Mild exogenous inflammation blunts neural signatures of bounded evidence accumulation and reward prediction error processing in healthy male participants.

BACKGROUND: Altered neural haemodynamic activity during decision making and learning has been linked to the effects of inflammation on mood and motivated behaviours. So far, it has been reported that blunted mesolimbic dopamine reward signals are associated with inflammation-induced anhedonia and apathy. Nonetheless, it is still unclear whether inflammation impacts neural activity underpinning decision dynamics. The process of decision making involves integration of noisy evidence from the environment until a critical threshold of evidence is reached. There is growing empirical evidence that such process, which is usually referred to as bounded accumulation of decision evidence, is affected in the context of mental illness. METHODS: In a randomised, placebo-controlled, crossover study, 19 healthy male participants were allocated to placebo and typhoid vaccination. Three to four hours post-injection, participants performed a probabilistic reversal-learning task during functional magnetic resonance imaging. To capture the hidden neurocognitive operations underpinning decision-making, we devised a hybrid sequential sampling and reinforcement learning computational model. We conducted whole brain analyses informed by the modelling results to investigate the effects of inflammation on the efficiency of decision dynamics and reward learning. RESULTS: We found that during the decision phase of the task, typhoid vaccination attenuated neural signatures of bounded evidence accumulation in the dorsomedial prefrontal cortex, only for decisions requiring short integration time. Consistent with prior work, we showed that, in the outcome phase, mild acute inflammation blunted the reward prediction error in the bilateral ventral striatum and amygdala. CONCLUSIONS: Our study extends current insights into the effects of inflammation on the neural mechanisms of decision making and shows that exogenous inflammation alters neural activity indexing efficiency of evidence integration, as a function of choice discriminability. Moreover, we replicate previous findings that inflammation blunts striatal reward prediction error signals.

Intact Mismatch Negativity Responses in Clinical High Risk for Psychosis and First-Episode Psychosis: Evidence From Source-Reconstructed Event-Related Fields and Time-Frequency Data.

BACKGROUND: This study examined whether mismatch negativity (MMN) responses are impaired in participants at clinical high risk for psychosis (CHR-P) and patients with first-episode psychosis (FEP) and whether MMN deficits predict clinical outcomes in CHR-Ps. METHODS: Magnetoencephalography data were collected during a duration-deviant MMN paradigm for a group of 116 CHR-P participants, 33 FEP patients (15 antipsychotic-naïve), clinical high risk negative group (n = 38) with substance abuse and affective disorder, and 49 healthy control participants. Analysis of group differences of source-reconstructed event-related fields as well as time-frequency and intertrial phase coherence focused on the bilateral Heschl's gyri and bilateral superior temporal gyri. RESULTS: Significant magnetic MMN responses were found across participants in the bilateral Heschl's gyri and bilateral superior temporal gyri. However, MMN amplitude as well as time-frequency and intertrial phase coherence responses were intact in CHR-P participants and FEP patients compared with healthy control participants. Furthermore, MMN deficits were not related to persistent attenuated psychotic symptoms or transitions to psychosis in CHR-P participants. CONCLUSIONS: Our data suggest that magnetic MMN responses in magnetoencephalography data are not impaired in early-stage psychosis and may not predict clinical outcomes in CHR-P participants.

Functional Magnetic Resonance Imaging Signatures of Pavlovian and Instrumental Valuation Systems during a Modified Orthogonalized Go/No-go Task.

Motivational (i.e., Pavlovian) values interfere with instrumental responding and can lead to suboptimal decision-making. In humans, task-based neuroimaging studies have only recently started illuminating the functional neuroanatomy of Pavlovian biasing of instrumental control. To provide a mechanistic understanding of the neural dynamics underlying the Pavlovian and instrumental valuation systems, analysis of neuroimaging data has been informed by computational modeling of conditioned behavior. Nonetheless, because of collinearities in Pavlovian and instrumental predictions, previous research failed to tease out hemodynamic activity that is parametrically and dynamically modulated by coexistent Pavlovian and instrumental value expectations. Moreover, neural correlates of Pavlovian to instrumental transfer effects have so far only been identified in extinction (i.e., in the absence of learning). In this study, we devised a modified version of the orthogonalized go/no-go paradigm, which introduced Pavlovian-only catch trials to better disambiguate trial-by-trial Pavlovian and instrumental predictions in both sexes. We found that hemodynamic activity in the ventromedial pFC covaried uniquely with the model-derived Pavlovian value expectations. Notably, modulation of neural activity encoding for instrumental predictions in the supplementary motor cortex was linked to successful action selection in conflict conditions. Furthermore, hemodynamic activity in regions pertaining to the limbic system and medial pFC was correlated with synergistic Pavlovian and instrumental predictions and improved conditioned behavior during congruent trials. Altogether, our results provide new insights into the functional neuroanatomy of decision-making and corroborate the validity of our variant of the orthogonalized go/no-go task as a behavioral assay of the Pavlovian and instrumental valuation systems.

A systematic review of factors prolonging or reducing the duration of untreated psychosis for people with psychosis in low- and middle-income countries.

AIM: This review aims to identify factors that may prolong or reduce the duration of untreated psychosis for people with psychosis in low- and middle-income countries. METHODS: Electronic searches of six databases were conducted, to find studies from low- and middle-income countries on people with psychotic disorders provided they statistically measured an association between factors that may prolong or reduce the duration of untreated psychosis. Studies were critically appraised and a narrative synthesis exploring differences between and within studies is presented. A socio-ecological model is used to convey the main findings. RESULTS: Thirty studies of 16 473 participants in total were included in this review. Taken together participants were 51.5% male and 48.5% female. Various factors potentially associated with longer duration of untreated psychosis for people with psychosis in low- and middle-income countries were found. Examples of these factors are an insidious mode of onset, greater family stigma and low social class. Other factors, such as marital status, educational level, diagnostic type, predominant symptoms and employment status, yielded inconsistent results. CONCLUSIONS: The methodological quality of the included studies limits the conclusions of this review. The results indicate an urgent need for further high-quality research in these countries. The socio-ecological model is a helpful framework for clinicians, scholars, and decision-makers to conceptualize factors that may affect the duration of untreated psychosis, highlight gaps in the literature as well as reflect on potential prevention strategies that may ultimately support early intervention services for people with psychosis in developing countries.

Thalamo-cortical circuits during sensory attenuation in emerging psychosis: a combined magnetoencephalography and dynamic causal modelling study.

Evidence suggests that schizophrenia (ScZ) involves impairments in sensory attenuation. It is currently unclear, however, whether such deficits are present during early-stage psychosis as well as the underlying network and the potential as a biomarker. To address these questions, Magnetoencephalography (MEG) was used in combination with computational modeling to examine M100 responses that involved a "passive" condition during which tones were binaurally presented, while in an "active" condition participants were asked to generate a tone via a button press. MEG data were obtained from 109 clinical high-risk for psychosis (CHR-P) participants, 23 people with a first-episode psychosis (FEP), and 48 healthy controls (HC). M100 responses at sensor and source level in the left and right thalamus (THA), Heschl's gyrus (HES), superior temporal gyrus (STG) and right inferior parietal cortex (IPL) were examined and dynamic causal modeling (DCM) was performed. Furthermore, the relationship between sensory attenuation and persistence of attenuated psychotic symptoms (APS) and transition to psychosis was investigated in CHR-P participants. Sensory attenuation was impaired in left HES, left STG and left THA in FEP patients, while in the CHR-P group deficits were observed only in right HES. DCM results revealed that CHR-P participants showed reduced top-down modulation from the right IPL to the right HES. Importantly, deficits in sensory attenuation did not predict clinical outcomes in the CHR-P group. Our results show that early-stage psychosis involves impaired sensory attenuation in auditory and thalamic regions but may not predict clinical outcomes in CHR-P participants.

Structural and Functional Brain Changes in Acute Takotsubo Syndrome.

BACKGROUND: Takotsubo syndrome mimics an acute myocardial infarction, typically in the aftermath of mental or physical stress. OBJECTIVES: The mechanism by which emotional processing in the context of stress leads to significant cardiac injury is poorly understood, so a full exploration of brain structure and function in takotsubo syndrome patients merits investigation. METHODS: Twenty-five acute (<5 days) takotsubo patients and 25 control subjects were recruited into this observational cross-sectional study. Surface-based morphometry was carried out on magnetic resonance imaging (MRI) brain scans to extract cortical morphology based on volume, thickness, and surface area with the use of Freesurfer. Cortical morphology general linear models were corrected for age, sex, photoperiod, and total brain volume. Resting-state functional MRI and diffusion tensor tractography images were preprocessed and analyzed with the use of the Functional Magnetic Resonance Imaging of the Brain Diffusion Toolbox and Functional Connectivity Toolbox. RESULTS: There was significantly smaller total white matter and subcortical gray matter volumes in takotsubo (P < 0.001), with smaller total brain surface area but increased total cortical thickness (both P < 0.001). Individual gray matter regions (hippocampus and others) were significantly smaller in takotsubo (P < 0.001); only thalamus and insula were larger (P < 0.001). There was significant hyperfunctional and hypofunctional connectivity in multiple areas, including thalamus-amygdala-insula and basal ganglia (P < 0.05). All structural tractography connections were increased in takotsubo (P < 0.05). CONCLUSIONS: The authors showed smaller gray and white matter volumes driven by smaller cortical surface area, but increased cortical thickness and structural tractography connections with bidirectional changes in functional connectivity linked to emotion, language, reasoning, perception, and autonomic control. These are interventional targets in takotsubo patients' rehabilitation.

Hippocampal structural alterations in early-stage psychosis: Specificity and relationship to clinical outcomes.

Hippocampal dysfunctions are a core feature of schizophrenia, but conflicting evidence exists whether volumetric and morphological changes are present in early-stage psychosis and to what extent these deficits are related to clinical trajectories. In this study, we recruited individuals at clinical high risk for psychosis (CHR-P) (n = 108), patients with a first episode of psychosis (FEP) (n = 37), healthy controls (HC) (n = 70) as well as a psychiatric control group with substance abuse and affective disorders (CHR-N: n = 38). MRI-data at baseline were obtained and volumetric as well as vertex analyses of the hippocampus were carried out. Moreover, volumetric changes were examined in the amygdala, caudate, nucleus accumbens, pallidum, putamen and thalamus. In addition, we obtained follow-up functional and symptomatic assessments in CHR-P individuals to examine the question whether anatomical deficits at baseline predicted clinical trajectories. Our results show that the hippocampus is the only structure showing significant volumetric decrease in early-stage psychosis, with FEPs showing significantly smaller hippocampal volumes bilaterally alongside widespread shape changes in the vertex analysis. For the CHR-P group, volumetric decreases were confined to the left hippocampus. However, hippocampal alterations in the CHR-P group were not robustly associated with clinical outcomes, including the persistence of attenuated psychotic symptoms and functional trajectories. Accordingly, our findings highlight that dysfunctions in hippocampal anatomy are an important feature of early-stage psychosis which may, however, not be related to clinical outcomes in CHR-P participants.

Five points to consider when reading a translational machine-learning paper.

Machine-learning techniques are used in this BJPsych special issue on precision medicine in attempts to create statistical models that make clinically relevant predictions for individual patients. In this primer, we outline five key points that are helpful for a new reader to consider in order to engage with the field and evaluate the literature. These points include the consideration of why we are interested in new statistical approaches, how they may produce individualised predictions, what caveats need to be kept in-mind and why the interest and engagment of clinicians and clinical researchers is critical to successful model development and implementation. We hope that the following primer will provide shared understanding to encourage dialogue between clinical and methodological fields.

Development and Validation of a Nonremission Risk Prediction Model in First-Episode Psychosis: An Analysis of 2 Longitudinal Studies.

Psychosis is a major mental illness with first onset in young adults. The prognosis is poor in around half of the people affected, and difficult to predict. The few tools available to predict prognosis have major weaknesses which limit their use in clinical practice. We aimed to develop and validate a risk prediction model of symptom nonremission in first-episode psychosis. Our development cohort consisted of 1027 patients with first-episode psychosis recruited between 2005 and 2010 from 14 early intervention services across the National Health Service in England. Our validation cohort consisted of 399 patients with first-episode psychosis recruited between 2006 and 2009 from a further 11 English early intervention services. The one-year nonremission rate was 52% and 54% in the development and validation cohorts, respectively. Multivariable logistic regression was used to develop a risk prediction model for nonremission, which was externally validated. The prediction model showed good discrimination C-statistic of 0.73 (0.71, 0.75) and adequate calibration with intercept alpha of 0.12 (0.02, 0.22) and slope beta of 0.98 (0.85, 1.11). Our model improved the net-benefit by 15% at a risk threshold of 50% compared to the strategy of treating all, equivalent to 15 more detected nonremitted first-episode psychosis individuals per 100 without incorrectly classifying remitted cases. Once prospectively validated, our first episode psychosis prediction model could help identify patients at increased risk of nonremission at initial clinical contact.

Anti-NMDAR encephalitis for psychiatrists: the essentials.

Anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis predominantly presents with psychiatric symptoms. Psychiatrists need to be alert to this diagnostic possibility, especially in female adolescents and young adults, as exemplified by the real (de-identified) case outlined below. Earlier diagnosis and immunotherapy improves long-term outcomes. Collaboration with neurology colleagues is essential for optimal care. 'Red flags' for autoimmune encephalitis and 'diagnostic clues' act as helpful aide memoires for this uncommon condition. The gold standard for testing is the detection of NMDAR antibodies in cerebrospinal fluid, but serum can be tested as a more accessible (but less reliable) preliminary step. The results of routine blood tests, magnetic resonance imaging of the head and electroencephalograms can be normal or show non-specific changes. Diagnostic criteria exist to define probable and definite cases. Immunotherapy for anti-NMDAR encephalitis is effective for many patients, but recovery is prolonged and relapses can occur.

Long range temporal correlations (LRTCs) in MEG-data during emerging psychosis: Relationship to symptoms, medication-status and clinical trajectory.

Long-Range Temporal Correlations (LRTCs) index the capacity of the brain to optimally process information. Previous research has shown that patients with chronic schizophrenia present altered LRTCs at alpha and beta oscillations. However, it is currently unclear at which stage of schizophrenia aberrant LRTCs emerge. To address this question, we investigated LRTCs in resting-state magnetoencephalographic (MEG) recordings obtained from patients with affective disorders and substance abuse (clinically at low-risk of psychosis, CHR-N), patients at clinical high-risk of psychosis (CHR-P) (n = 115), as well as patients with a first episode (FEP) (n = 25). Matched healthy controls (n = 47) served as comparison group. LRTCs were obtained for frequencies from 4 to 40 Hz and correlated with clinical and neuropsychological data. In addition, we examined the relationship between LRTCs and transition to psychosis in CHR-P participants, and the relationship between LRTC and antipsychotic medication in FEP participants. Our results show that participants from the clinical groups have similar LRTCs to controls. In addition, LRTCs did not correlate with clinical and neurocognitive variables across participants nor did LRTCs predict transition to psychosis. Therefore, impaired LRTCs do not reflect a feature in the clinical trajectory of psychosis. Nevertheless, reduced LRTCs in the beta-band over posterior sensors of medicated FEP participants indicate that altered LRTCs may appear at the onset of the illness. Future studies are needed to elucidate the role of anti-psychotic medication in altered LRTCs.

40-Hz Auditory Steady-State Responses Characterize Circuit Dysfunctions and Predict Clinical Outcomes in Clinical High-Risk for Psychosis Participants: A Magnetoencephalography Study.

BACKGROUND: This study aimed to examine whether 40-Hz auditory steady-state responses (ASSRs) are impaired in participants at clinical high-risk for psychosis (CHR-P) and predict clinical outcomes. METHODS: Magnetoencephalography data were collected during a 40-Hz ASSR paradigm for a group of 116 CHR-P participants, 33 patients with first-episode psychosis (15 antipsychotic-naïve), a psychosis risk-negative group (n = 38), and 49 healthy control subjects. Analysis of group differences of 40-Hz intertrial phase coherence and 40-Hz amplitude focused on right Heschl's gyrus, superior temporal gyrus, hippocampus, and thalamus after establishing significant activations during 40-Hz ASSR stimulation. Linear regression and linear discriminant analyses were used to predict clinical outcomes in CHR-P participants, including transition to psychosis and persistence of attenuated psychotic symptoms (APSs). RESULTS: CHR-P participants and patients with first-episode psychosis were impaired in 40-Hz amplitude in the right thalamus and hippocampus. In addition, patients with first-episode psychosis were impaired in 40-Hz amplitude in the right Heschl's gyrus, and CHR-P participants in 40-Hz intertrial phase coherence in the right Heschl's gyrus. The 40-Hz ASSR deficits were pronounced in CHR-P participants who later transitioned to psychosis (n = 13) or showed persistent APSs (n = 34). Importantly, both APS persistence and transition to psychosis were predicted by 40-Hz ASSR impairments, with ASSR activity in the right hippocampus, superior temporal gyrus, and middle temporal gyrus correctly classifying 69.2% individuals with nonpersistent APSs and 73.5% individuals with persistent APSs (area under the curve = 0.842), and right thalamus 40-Hz activity correctly classifying 76.9% transitioned and 53.6% nontransitioned CHR-P participants (area under the curve = 0.695). CONCLUSIONS: Our data indicate that deficits in gamma-band entrainment in the primary auditory cortex and subcortical areas constitute a potential biomarker for predicting clinical outcomes in CHR-P participants.

Mild inflammation causes a reduction in resting-state amplitude of low-frequency fluctuation in healthy adult males.

Systemic inflammation has been associated with negative mood states and human sickness behaviour. Previous studies have shown an association between systemic inflammation and changes in task-related blood-oxygen-level-dependent activity and functional connectivity within large-scale networks. However, no study has examined the effect of inflammation on the magnitude of blood-oxygen-level-dependent low-frequency fluctuations at rest. We used a double-blind placebo-controlled crossover design to randomise 20 male subjects (aged 20-50 years) to receive either a Salmonella typhi vaccine or a placebo saline injection at two separate sessions. All participants underwent a resting-state functional magnetic resonance scan and a measure of inflammation (interleukin 6) and mood (Profile of Mood States) 3 h after injection. We compared the whole brain amplitude of low-frequency fluctuations between the vaccine and placebo conditions using a repeated measures design. Vaccine condition was associated with greater interleukin 6 levels (p < 0.001). Vaccine condition was also associated with lower amplitude of low-frequency fluctuations in the right and left frontal pole, superior frontal gyrus, paracingulate gyrus (Cluster 1) and the right mid and inferior frontal gyrus (Cluster 2) (p < 0.001, false discovery rate corrected). Lower amplitude of low-frequency fluctuations pertaining to first cluster correlated with greater total Profile of Mood States score (worse mood) (r = -0.38; p = 0.04). These results imply possible excitation/inhibition imbalance mechanisms during inflammation that may be a relevant target in psychiatric disease, especially mood disorders.

Mild Inflammation in Healthy Males Induces Fatigue Mediated by Changes in Effective Connectivity Within the Insula.

BACKGROUND: Systemic inflammation is associated with sickness behaviors such as low mood and fatigue. Activity patterns within the insula are suggested to coordinate these behaviors but have not been modeled. We hypothesized that mild systemic inflammation would result in changes in effective connectivity between the viscerosensory and the visceromotor regions of the insula. METHODS: We used a double-blind, crossover design to randomize 20 male subjects to receive either a Salmonella typhi vaccine or a placebo saline injection at two separate sessions. All participants underwent a resting-state functional magnetic resonance scan 3 hours after injection. We determined behavioral and inflammatory changes, using the Profile of Mood States questionnaire and interleukin-6 levels. We extracted effective connectivity matrices between bilateral mid/posterior (viscerosensory) and anterior (visceromotor) insular cortices using spectral dynamic causal modeling. We applied parametric empirical Bayes and mediation analysis to determine a vaccination effect on effective connectivity and whether this mediated behavioral changes. RESULTS: The vaccine condition was associated with greater interleukin-6 levels and greater fatigue 3 hours after the injection. Activity within the right mid/posterior insula increased the activity within the bilateral anterior insular regions. This connectivity was augmented by vaccination over a 99% posterior confidence threshold. The right mid/posterior insula-to-left anterior insula connectivity was significantly associated with fatigue and mediated the association between inflammation and increased fatigue scores. CONCLUSIONS: These results demonstrate that increased effective connectivity between specific nodes of the insula can model and mediate the association between inflammation and fatigue in males.

Association of Magnetoencephalographically Measured High-Frequency Oscillations in Visual Cortex With Circuit Dysfunctions in Local and Large-scale Networks During Emerging Psychosis.

Importance: Psychotic disorders are characterized by impairments in neural oscillations, but the nature of the deficit, the trajectory across illness stages, and functional relevance remain unclear. Objectives: To examine whether changes in spectral power, phase locking, and functional connectivity in visual cortex are present during emerging psychosis and whether these abnormalities are associated with clinical outcomes. Design, Setting, and Participants: In this cross-sectional study, participants meeting clinical high-risk criteria for psychosis, participants with first-episode psychosis, participants with affective disorders and substance abuse, and a group of control participants were recruited. Participants underwent measurements with magnetoencephalography and magnetic resonance imaging. Data analysis was carried out between 2018 and 2019. Main Outcomes and Measures: Magnetoencephalographical activity was examined in the 1- to 90-Hz frequency range in combination with source reconstruction during a visual grating task. Event-related fields, power modulation, intertrial phase consistency, and connectivity measures in visual and frontal cortices were associated with neuropsychological scores, psychosocial functioning, and clinical symptoms as well as persistence of subthreshold psychotic symptoms at 12 months. Results: The study participants included those meeting clinical high-risk criteria for psychosis (n = 119; mean [SD] age, 22 [4.4] years; 32 men), 26 patients with first-episode psychosis (mean [SD] age, 24 [4.2] years; 16 men), 38 participants with affective disorders and substance abuse (mean [SD] age, 23 [4.7] years; 11 men), and 49 control participants (mean age [SD], 23 [3.6] years; 16 men). Clinical high-risk participants and patients with first-episode psychosis were characterized by reduced phase consistency of ß/γ-band oscillations in visual cortex (d = 0.63/d = 0.93). Moreover, the first-episode psychosis group was also characterized by reduced occipital γ-band power (d = 1.14) and altered visual cortex connectivity (d = 0.74-0.84). Impaired fronto-occipital connectivity was present in both clinical high-risk participants (d = 0.54) and patients with first-episode psychosis (d = 0.84). Importantly, reductions in intertrial phase coherence predicted persistence of subthreshold psychosis in clinical high-risk participants (receiver operating characteristic area under curve = 0.728; 95% CI, 0.612-0.841; P = .001). Conclusions and Relevance: High-frequency oscillations are impaired in the visual cortex during emerging psychosis and may be linked to behavioral and clinical impairments. Impaired phase consistency of γ-band oscillations was also associated with the persistence of subthreshold psychosis, suggesting that magnetoencephalographical measured neural oscillations could constitute a biomarker for clinical staging of emerging psychosis.

Predicting one-year outcome in first episode psychosis using machine learning.

BACKGROUND: Early illness course correlates with long-term outcome in psychosis. Accurate prediction could allow more focused intervention. Earlier intervention corresponds to significantly better symptomatic and functional outcomes. Our study objective is to use routinely collected baseline demographic and clinical characteristics to predict employment, education or training (EET) status, and symptom remission in patients with first episode psychosis (FEP) at one-year. METHODS AND FINDINGS: 83 FEP patients were recruited from National Health Service (NHS) Glasgow between 2011 and 2014 to a 24-month prospective cohort study with regular assessment of demographic and psychometric measures. An external independent cohort of 79 FEP patients were recruited from NHS Glasgow and Edinburgh during a 12-month study between 2006 and 2009. Elastic net regularised logistic regression models were built to predict binary EET status, period and point remission outcomes at one-year on 83 Glasgow patients (training dataset). Models were externally validated on an independent dataset of 79 patients from Glasgow and Edinburgh (validation dataset). Only baseline predictors shared across both cohorts were made available for model training and validation. After excluding participants with missing outcomes, models were built on the training dataset for EET status, period and point remission outcomes and externally validated on the validation dataset. Models predicted EET status, period and point remission with receiver operating curve (ROC) area under the curve (AUC) performances of 0.876 (95%CI: 0.864, 0.887), 0.630 (95%CI: 0.612, 0.647) and 0.652 (95%CI: 0.635, 0.670) respectively. Positive predictors of EET included baseline EET and living with spouse/children. Negative predictors included higher PANSS suspiciousness, hostility and delusions scores. Positive predictors for symptom remission included living with spouse/children, and affective symptoms on the Positive and Negative Syndrome Scale (PANSS). Negative predictors of remission included passive social withdrawal symptoms on PANSS. A key limitation of this study is the small sample size (n) relative to the number of predictors (p), whereby p approaches n. The use of elastic net regularised regression rather than ordinary least squares regression helped circumvent this difficulty. Further, we did not have information for biological and additional social variables, such as nicotine dependence, which observational studies have linked to outcomes in psychosis. CONCLUSIONS AND RELEVANCE: Using advanced statistical machine learning techniques, we provide the first externally validated evidence, in a temporally and geographically independent cohort, for the ability to predict one-year EET status and symptom remission in individual FEP patients.

Development and validation of multivariable prediction models of remission, recovery, and quality of life outcomes in people with first episode psychosis: a machine learning approach.

BACKGROUND: Outcomes for people with first-episode psychosis are highly heterogeneous. Few reliable validated methods are available to predict the outcome for individual patients in the first clinical contact. In this study, we aimed to build multivariable prediction models of 1-year remission and recovery outcomes using baseline clinical variables in people with first-episode psychosis. METHODS: In this machine learning approach, we applied supervised machine learning, using regularised regression and nested leave-one-site-out cross-validation, to baseline clinical data from the English Evaluating the Development and Impact of Early Intervention Services (EDEN) study (n=1027), to develop and internally validate prediction models at 1-year follow-up. We assessed four binary outcomes that were recorded at 1 year: symptom remission, social recovery, vocational recovery, and quality of life (QoL). We externally validated the prediction models by selecting from the top predictor variables identified in the internal validation models the variables shared with the external validation datasets comprised of two Scottish longitudinal cohort studies (n=162) and the OPUS trial, a randomised controlled trial of specialised assertive intervention versus standard treatment (n=578). FINDINGS: The performance of prediction models was robust for the four 1-year outcomes of symptom remission (area under the receiver operating characteristic curve [AUC] 0·703, 95% CI 0·664-0·742), social recovery (0·731, 0·697-0·765), vocational recovery (0·736, 0·702-0·771), and QoL (0·704, 0·667-0·742; p<0·0001 for all outcomes), on internal validation. We externally validated the outcomes of symptom remission (AUC 0·680, 95% CI 0·587-0·773), vocational recovery (0·867, 0·805-0·930), and QoL (0·679, 0·522-0·836) in the Scottish datasets, and symptom remission (0·616, 0·553-0·679), social recovery (0·573, 0·504-0·643), vocational recovery (0·660, 0·610-0·710), and QoL (0·556, 0·481-0·631) in the OPUS dataset. INTERPRETATION: In our machine learning analysis, we showed that prediction models can reliably and prospectively identify poor remission and recovery outcomes at 1 year for patients with first-episode psychosis using baseline clinical variables at first clinical contact. FUNDING: Lundbeck Foundation.

Dysregulated Brain Dynamics in a Triple-Network Saliency Model of Schizophrenia and Its Relation to Psychosis.

BACKGROUND: Schizophrenia is a highly disabling psychiatric disorder characterized by a range of positive "psychosis" symptoms. However, the neurobiology of psychosis and associated systems-level disruptions in the brain remain poorly understood. Here, we test an aberrant saliency model of psychosis, which posits that dysregulated dynamic cross-network interactions among the salience network (SN), central executive network, and default mode network contribute to positive symptoms in patients with schizophrenia. METHODS: Using task-free functional magnetic resonance imaging data from two independent cohorts, we examined 1) dynamic time-varying cross-network interactions among the SN, central executive network, and default mode network in 130 patients with schizophrenia versus well-matched control subjects; 2) accuracy of a saliency model-based classifier for distinguishing dynamic brain network interactions in patients versus control subjects; and 3) the relation between SN-centered network dynamics and clinical symptoms. RESULTS: In both cohorts, we found that dynamic SN-centered cross-network interactions were significantly reduced, less persistent, and more variable in patients with schizophrenia compared with control subjects. Multivariate classification analysis identified dynamic SN-centered cross-network interaction patterns as factors that distinguish patients from control subjects, with accuracies of 78% and 80% in the two cohorts, respectively. Crucially, in both cohorts, dynamic time-varying measures of SN-centered cross-network interactions were correlated with positive, but not negative, symptoms. CONCLUSIONS: Aberrations in time-varying engagement of the SN with the central executive network and default mode network is a clinically relevant neurobiological signature of psychosis in schizophrenia. Our findings provide strong evidence for dysregulated brain dynamics in a triple-network saliency model of schizophrenia and inform theoretically motivated systems neuroscience approaches for characterizing aberrant brain dynamics associated with psychosis.