2026 WFSBP指南：精神病的预防

ABSTRACT

Objective: To prepare evidence-based guidelines on psychosis prevention.

Methods: We reviewed evidence on risk factors for/age at onset of psychosis, tools to assess clinical high-risk of psychosis (CHR-P), transition rates, risk calculation/ethical considerations around risk communication, CHR-P biological/clinical correlates, efficacy/cost-effectiveness of interventions/psychosis prevention services. The World Federation of Societies of Biological Psychiatry framework was used to grade evidence regarding interventions/services, elaborating guidelines with evidence-/consensus-based clinical recommendations to prevent psychosis in CHR-P subjects.

Results: At the service organisation level, (i) psychosis indicated prevention services might be implemented in close collaboration with early intervention services for psychosis to minimise duration of untreated illness, (ii) intake age criteria should be between 14 to 35, (iii) services should allow access to persons with cannabis use disorder. At the assessment and risk communication level, in clinical settings: (iv) staff in mental health services should be trained in administering/rating CHR-P assessment tools, (v) administer them, (vi) be trained in using/ interpreting risk calculators, and (vii) in communicating risk, (viii) only use validated risk calculators, keeping a human-to-human interaction. Also, (ix) prevention services should assess comorbid mental disorders. At the intervention level: (x) staff should offer treatment for abstinence from cannabis, (xi) offer evidence-based treatment for comorbid mental disorders, and (xii) offer treatment for CHR-P based on patient preference, following the 'first do no harm' principle.

Conclusions: Prevention services should be implemented, including interventions for cannabis use, reducing the duration of untreated psychosis, and treating comorbid mental disorders.

ARTICLE HISTORY

Received 10 October 2025

Revised 17 March 2026

Accepted 19 March 2026

KEYWORDS

Clinical guidelines; psychosis; prevention; clinical high risk for psychosis; schizophrenia

1. Background

Psychotic disorders, such as schizophrenia, are associated with considerable burden (Solmi, Seitidis, et al. 2023) and reduced life expectancy compared with the general population (Correll et al. 2022) due to suicide, but most frequently as a result of co-occurring physical conditions (Stubbs et al. 2016; Correll et al. 2017). The reduced life expectancy in people with schizophrenia is aggravated by the low quality of physical healthcare that people with schizophrenia receive (Solmi, Firth, et al. 2020; Solmi, Fiedorowicz, et al. 2021). In the absence of early interventions, schizophrenia often has a chronic course, severely impacting patients' functioning and their families' environment (Kahn et al. 2015). Effective treatments for schizophrenia are available, with antipsychotics being the cornerstone of treatment at all stages of the disease (Solmi et al. 2017; Correll, Rubio, et al. 2018; McCutcheon et al. 2025). Well-evidenced psychosocial interventions should augment pharmacological treatment, in particular within early intervention services (Correll, Galling, et al. 2018). These include family interventions (for the early stages of schizophrenia), as well as cognitive behavioural therapy (CBT), cognitive remediation therapy, and individual placement and support (Solmi, Croatto, et al. 2023). While considered a new treatment modality, brain stimulation is also effective for schizophrenia, yet they are not first-line treatments (Rosson et al. 2022). Lifestyle interventions are increasingly becoming accepted and recommended (Gurusamy et al. 2018; Stubbs et al. 2018; Firth et al. 2020). Despite

the several treatment options, none of these are disease-modifying. Hence, interventions that prevent or delay the onset of schizophrenia are needed, especially considering that the duration of untreated psychosis has been associated with poorer medium and long-term outcomes (Howes et al. 2021).

While prevention has been largely implemented in other branches of medicine, it is only recently being applied to mental health care (Arango et al. 2018; Kotlicka-Antczak et al. 2020; Fusar-Poli, Correll, et al. 2021; Estradé et al. 2022). Within mental health, many studies have focused on the indicated prevention of psychosis, mainly in subjects presenting psychotic symptoms that do not meet the intensity or duration criteria for a Diagnostic and Statistical Manual of Mental Disorders (DSM)- or International Classification of Diseases (ICD)- defined psychotic disorder. The presence of subthreshold psychotic symptoms has been defined as an ultra-high or clinical high-risk (UHR/CHR) state, and has been introduced more than 25 years ago in Melbourne, Australia (Yung et al. 2005). Currently, evidence on the possibility of preventing psychosis with specific interventions has been conflicting (Mei et al. 2021; Fusar-Poli, Radua, et al. 2022), and has mainly focused on subjects meeting criteria for CHR of psychosis (CHR-P), who comprise the majority of those ultimately developing a psychotic disorder (Shah et al. 2017; Benrimoh et al. 2024).

The following guidelines provide a comprehensive overview on risk factors for psychosis, the definition and tools to assess CHR-P status, its biological

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

3

correlates and comorbid mental disorders, the literature around risk of transition from CHR-P to psychosis and around additional outcomes in CHR-P. We also reviewed literature around individual randomised controlled trials (RCTs) testing interventions to prevent psychosis, (network) meta-analyses pooling results of these RCTs, and finally observational or simulation studies reporting on the clinical and cost-effectiveness outcomes of psychosis prevention services. Recommendations for the indicated prevention of psychosis in subjects meeting CHR-P criteria were made, based on consensus, on the evaluation of RCTs and evidence regarding psychosis prevention services according to the World Federation of Societies of Biological Psychiatry (WFSBP) guidance for preparing clinical guidelines (Hasan et al. 2019).

2. Methods

We summarised available evidence adopting several strategies including meta-umbrella, umbrella, systematic reviews, or relying on previous systematic evidence synthesis recently completed, on a set of a-priori selected clinically relevant areas related to prevention of psychosis. Starting from the literature knowledge of authors, we identified relevant reviews, and updated the search to 19/09/2024 if the most updated evidence synthesis publication was published before September 2022 (more than two years old), prioritising meta-umbrella over umbrella and systematic reviews, and following the same methodology of the most updated review (e.g. an updated umbrella review for

(meta-) umbrella reviews, an updated systematic review for systematic reviews). We chose September 2022, i.e. two years, for feasibility considerations given the breadth of this evidence synthesis. Where no previous evidence synthesis was available, we conducted one, with three exceptions. First, we did not update reviews that included over 100 epidemiologic studies, since we anticipated that adding few studies would have not been likely to materially change results. Second, we did not conduct systematic reviews to identify the most frequently used psychometric tools to assess CHR-P, given the robust expertise of the authors on this subject (i.e. the tools are already known). Third, we did not conduct a systematic review for ethical considerations around the risk in assessing and communicating risk in psychosis prediction, since the subject would not easily lend itself to quantitative evidence syntheses. Instead, we leveraged the report of a recent workshop completed by experts in the area, well representing the multifaceted set of considerations around ethics and risk prediction. We then graded evidence following the WFSBP guidelines for clinical guidelines development (Hasan et al. 2019).

The a-priori list of topics, search keys, study selection flows, as well as methodological framework used for guideline preparation are detailed in the supplementary material, and in Table 1. The approach followed to review literature on each specific area is summarised below, including the criteria grading credibility of the evidence on risk factors (Table 2). Finally, each recommendation reached a consensus among co-authors as detailed below.

Table 1. World Federation of Societies of Biological Psychiatry grading system, and grades of recommendations.

|Grades of recommendation are based on a synthesis of: i. Level of evidence (see Table 10), and ii. Acceptability (criteria for grading adapted and modified from (AWMF 2012; GRADE 2017), rated 'strong,' limited', and 'weak'. Acceptability criteria are factors considered are risk-benefit ratio (e.g. adverse effects, interactions), cost-benefit ratio, applicability in the target population, ethical and legal aspects, preferences of service users, practicability.The algorithm of evaluating the acceptability to develop GoRs from LoEs is detailed below, for positive and negative recommendations.|||||| |---|---|---|---|---|---| |Recommendations for using the intervention|Recommendations for using the intervention||||| |Strong|1|A' LoE and GOOD acceptability|Strong|1|Strong negative evidence (LoE -A)| |Limited|2|A' LoE and MODERATE acceptabilityOR' LoE and GOOD acceptability|Limited|2|Limited negative evidence (LoE -B)| |Weak|3|A' LoE and POOR acceptabilityOR' LoE and MODERATE/POOR acceptabilityOR' LoE and GOOD/MODERATE/|Weak|3|Weak negative evidence (LoE-C)| |No recommendation possible|4|Insufficient evidence (LoE D) to give recommendations|No recommendation possible|4|Insufficient evidence (LoE D) to give recommendations|

Legend. GoR: grade of recommendation; LoE: level of evidence.

4 M. SOLMI ET AL.

Table 2. Risk factors and biomarkers with significant association with psychosis reported in meta-analyses and umbrella reviews, and their credibility according to established criteria (loannidis 2009; Fusar-Poli and Radua 2018; Solmi, Dragoti, Croatto, Radua, Borgwardt, Carvalho, Demurtas, Mosina, Kurotschka, Shin, et al. 2021; Solmi, De Toffol, et al. 2023).

|Risk factor|k|Association odds ratio|Credibility| |---|---|---|---| |Individual medical, psychiatric, psychological factors|||| |Ultra-high-risk state for psychosis|9|9.32 (4.91-17.32)|Convincing| |Cannabis use|18|1.71 (1.47-2.00)|Highly suggestive/Convincing| |Poor olfactory identification ability|55|5.26 (NA-NA)|Highly suggestive| |Trait anhedonia|44|4.41 (NA-NA)|Highly suggestive| |Low premorbid IQ|16|2.13 (NA-NA)|Highly suggestive| |Childhood social withdrawal|15|2.91 (NA-NA)|Suggestive| |Tobacco use|17|2.34 (1.65-3.33)|Suggestive| |Non-right handedness|41|1.58 (1.35-1.86)|Suggestive| |Neurological soft signs|8|27.59 (NA-NA)|Weak| |Poor olfactory memory ability|2|20 (NA-NA)|Weak| |Neuroticism|8|8.76 (NA-NA)|Weak| |Epilepsy|4|7.84 (2.87-21.40)|Weak| |Harm avoidance|7|5.92 (NA-NA)|Weak| |Poor self-directedness|7|5.88 (NA-NA)|Weak| |Heavy cannabis use|2|5.17 (3.64-7.36)|Weak| |Poor olfactory discrimination ability|8|5.00 (NA-NA)|Weak| |Poor extraversion|8|5.00 (NA-NA)|Weak| |Poor olfactory hedonic ability|10|4.00 (NA-NA)|Weak| |Psychotic-like experiences|4|3.84 (2.55-5.79)|Weak| |Poor conscientiousness|7|3.45 (NA-NA)|Weak| |Hearing impairment|3|3.15 (1.25-7.95)|Weak| |Poor olfactory detection ability|18|3.12 (NA-NA)|Weak| |High Self-transcendence|7|3.03 (NA-NA)|Weak| |Poor motor function pre-onset of psychosis|4|2.78 (NA-NA)|Weak| |Antisocial and externalising behaviour|3|2.39 (NA-NA)|Weak| |Poor agreeableness|6|2.38 (NA-NA)|Weak| |Delay in walking unsupported|5|2.37 (NA-NA)|Weak| |Poor cooperativeness|7|2.33 (NA-NA)|Weak| |Hypoalgesia|9|2.31 (NA-NA)|Weak| |Poor reward dependence|7|2.17 (NA-NA)|Weak| |Poor openness|7|2.04 (NA-NA)|Weak| |Delay in standing unsupported|4|1.58 (NA-NA)|Weak| |Poor persistence (trait of temperament)|7|1.54 (NA-NA)|Weak| |Traumatic brain injury|8|1.49 (1.09-2.05)|Weak| |Delay in sitting unsupported|4|1.41 (NA-NA)|Weak| |Total A-B ridge count|13|1.32 (NA-NA)|Weak| |Delay in holding head up|3|1.26 (NA-NA)|Weak| |Biomarkers|||| |High Serum C-reactive protein|15|2.96 (1.75-5.01)|Suggestive| |Low serum brain-derived neurotrophic factor|17|2.94 (1.61-5.55)|Suggestive| |High Serum interleukin 6|29|2.61 (1.55-4.40)|Suggestive| |High Plasma MDA|4|42.88 (3.15-583.91)|Weak| |Low plasma TAS|3|16.67 (6.25-50.00)|Weak| |High Serum IL-8|4|10.57 (1.81-61.82)|Weak| |High Serum malondialdehyde|17|9.60 (3.27-28.19)|Weak| |Low serum 25(OH)D|6|8.33 (2.78-25.00)|Weak| |Low serum NGF|8|7.14 (1.18-50.00)|Weak| |High Serum DPA|14|7.10 (3.42-14.75)|Weak| |High Serum S100B|19|5.81 (3.03-11.17)|Weak| |High Serum IL-1β|11|5.45 (2.09-14.19)|Weak| |Impaired glucose tolerance|7|4.55 (2.10-9.84)|Weak| |Low serum Zink|9|4.35 (1.33-14.29)|Weak| |High Serum prolactin|12|4.26 (1.92-9.45)|Weak| |High Serum sil-2R|9|3.55 (2.00-6.32)|Weak| |High Serum LA|16|3.45 (1.67-7.13)|Weak| |High Plasma TBARS|12|3.20 (1.51-6.78)|Weak| |High Serum glutamate|10|3.17 (1.46-6.90)|Weak| |High Serum AA|19|3.16 (1.84-5.43)|Weak| |High Serum TNF-α|24|2.80 (1.50-5.21)|Weak| |Low Serum ADH|11|2.78 (1.30-5.88)|Weak| |High Serum DHA|18|2.52 (1.19-5.35)|Weak| |High Serum IL-1RA|4|2.46 (1.07-5.63)|Weak| |High Serum TGF-β|6|2.23 (1.29-3.84)|Weak| |Low plasma uric acid|4|1.89 (1.13-3.12)|Weak| |Serum homocysteine (Per 5 mmol/l increase)|8|1.68 (1.25-2.25)|Weak| |Insulin resistance|8|1.68 (1.09-2.59)|Weak| |High Serum serine|19|1.66 (1.04-2.66)|Weak| |High Serum leptin|28|1.62 (1.14-2.30)|Weak| |High Serum morning cortisol|42|1.60 (1.16-2.21)|Weak|

(Continued)

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

5

Table 2. Continued.

|Risk factor|k|Association odds ratio|Credibility| |---|---|---|---| |Unsaturated fatty acids supplementation|9|0.59 (0.39-0.89)|Weak| |Pregnancy factors|||| |Famine|3|1.61 (1.51-1.71)|Convincing| |Any famine or nutritional deficiency|13|1.40 (1.17-1.68)|Suggestive| |Herpes simplex type 2 infection|5|1.35 (1.16-1.58)|Suggestive| |Diabetes in pregnancy|2|10.12 (1.84-55.72)|Weak| |Polyhydramnos|3|3.05 (1.15-8.06)|Weak| |Maternal stress|4|2.40 (1.15-5.01)|Weak| |Premature ruptured membranes|4|2.29 (1.38-3.80)|Weak| |Threatened premature delivery|2|2.05 (1.02-4.10)|Weak| |Pre pregnancy and pregnancy maternal obesity|4|1.99 (1.26-3.14)|Weak| |Uterine atony|3|1.93 (1.35-2.76)|Weak| |Suboptimal number of antenatal care visits|4|1.83 (1.42-2.36)|Weak| |Antepartum haemorrhage|14|1.63 (1.12-2.38)|Weak| |Foetal hypoxia|6|1.63 (1.11-2.40)|Weak| |Maternal hypertension|7|1.40 (1.10-1.78)|Weak| |Toxoplasma infection|4|1.30 (1.04-1.62)|Weak| |Unspecified maternal infection|9|1.27 (1.06-1.53)|Weak| |Maternal and paternal factors|||| |Maternal psychosis|6|7.61 (6.29-9.21)|Highly suggestive| |Any maternal psychopathology|9|4.60 (2.74-7.73)|Highly suggestive| |Any paternal psychopathology|5|2.73 (2.33-3.19)|Highly suggestive| |Three or more previous pregnancies|10|1.30 (1.16-1.45)|Suggestive| |Maternal age < 20 years old|22|1.17 (1.07-1.27)|Suggestive| |Parental severe mental illness|9|5.94 (2.99-11.79)|Weak| |Paternal communication deviance|19|5.80 (3.99-8.44)|Weak| |Paternal age > 45 years|4|2.36 (1.35-4.11)|Weak| |Maternal affective disorder|3|2.26 (1.09-4.70)|Weak| |Low paternal socio-economic status|9|1.30 (1.02-1.65)|Weak| |Paternal age > 34 years old|13|1.28 (1.06-1.55)|Weak| |Maternal age 30-34 years old|20|1.05 (1.01-1.09)|Weak| |Ethnicity or migration-related factors|||| |Black-Caribbean ethnicity in England|9|4.87 (3.96-6.00)|Convincing| |Ethnic minority in low ethnic density area|5|3.71 (2.47-5.58)|Highly suggestive| |Second-generation immigrant|26|1.68 (1.42-1.92)|Highly suggestive| |North-African immigrants in Europe|12|2.22 (1.58-3.12)|Suggestive| |Ethnic minority in high ethnic density area|5|2.11 (1.39-3.20)|Suggestive| |First-generation immigrant|42|2.10 (1.72-2.56)|Suggestive| |Black African ethnicity in England|4|4.72 (3.30-6.77)|Weak| |Asian ethnicity in England|6|2.83 (1.59-5.05)|Weak| |Other white ethnicity in England|3|2.62 (1.35-5.10)|Weak| |Mixed ethnicity in England|3|2.19 (1.08-4.44)|Weak| |Foetal growth and gestational age|||| |Birth weight < 2kg|14|1.84 (1.53-2.22)|Convincing| |Birth weight < 2.5kg|25|1.53 (1.31-1.78)|Convincing| |Small for gestational age|12|1.40 (1.25-1.57)|Convincing| |Birth weight 2.5-2.999kg|17|1.23 (1.15-1.31)|Convincing| |Minor physical anomalies|14|5.30 (NA-NA)|Highly suggestive| |Congenital malformation|10|2.31 (1.29-4.13)|Weak| |Small head circumference|2|1.41 (1.00-1.97)|Weak| |Premature birth|24|1.35 (1.12-1.62)|Weak| |Birth length < 49 cm|8|1.17 (1.05-1.32)|Weak| |Labour season, labour complications, and delivery complications|||| |Unspecific obstetric complications|21|1.52 (1.19-1.94)|Weak| |Winter (Nordic countries)|11|1.05 (1.03-1.08)|Suggestive| |Winter to spring (Nordic countries)|27|1.04 (1.02-1.06)|Suggestive| |Incubator or resuscitation|8|2.12 (1.29-3.47)|Weak| |Asphyxia state|9|1.93 (1.30-2.88)|Weak| |Definite obstetric complication*|9|1.83 (1.21-2.77)|Weak| |Individual infections|||| |Toxoplasma lgG|42|1.82 (1.51-2.18)|Suggestive| |Chlamydia psittaci|2|29.05 (8.91-94.96)|Weak| |Human endogenous retrovirus type W|5|19.78 (6.50-60.22)|Weak| |Chlamydia pneumoniae|3|6.02 (2.86-12.66)|Weak| |Herpes simplex type 2 infection|6|1.44 (1.14-1.81)|Weak| |Borna disease virus|21|1.36 (1.09-1.70)|Weak| |Individual or environmental stress|||| |Adult stressful events|13|3.11 (2.31-4.18)|Highly suggestive| |Childhood adversities|34|2.80 (2.34-3.34)|Highly suggestive| |Any psychological trauma|25|2.66 (1.99-3.56)|Highly suggestive| |Childhood trauma|23|2.49 (1.86-3.33)|Highly suggestive| |Urbanicity|8|2.39 (1.63-3.51)|Suggestive| |Sexual abuse|2|12.24 (3.92-38.17)|Weak|

(Continued)

6 M. SOLMI ET AL.

Table 2. Continued.

|Risk factor|k|Association odds ratio|Credibility| |---|---|---|---| |Adult negative life events|6|5.34 (3.84-7.43)|Weak| |Disadvantaged group|3|2.27 (1.21-4.27)|Weak| |Risk factors for transition to psychosis within subjects at clinical high risk for psychosis|||| |Attenuated positive symptoms|49|2.563 (NA - NA)|Highly suggestive| |Poor global functioning|49|1.69 (NA - NA)|Highly suggestive| |Negative symptoms|49|2.68 (NA - NA)|Suggestive| |Disorganised/cognitive symptoms|18|2.48 (NA - NA)|Weak| |General symptoms|21|2.27 (NA - NA)|Weak| |Unemployment|7|1.81 (1.31-2.5)|Weak| |Right handeddress|16|1.60 (1.04-2.46)|Weak| |Stress/trauma|11|1.15 (1.04-1.26)|Weak| |Male sex|66|1.18 (1.03-1.34)|Weak| |Total symptoms|29|1.74 (NA-NA)|Weak|

Legend. 25(OH)D, 25-Hydroxyvitamin D; AA, Arachidonic Acid; ADH, Vasopressin; DHA, Docosahexaenoic Acid; DPA, Docosapentaenoic Acid; IL-1β, Interleukin-1 Beta; IL-1RA, Interleukin-1 Receptor Antagonist; IL-8, Interleukin-8; LA, Linoleic Acid; MDA, Malondialdehyde; NGF, Nerve Growth Factor; S100B, S100 Calcium-Binding Protein B; $IL-2R$, Soluble Interleukin-2 Receptor, TAS, Total Antioxid Status; TBARS, Thiobarbituric Acid Reactive Substances; TGF-β, Transforming Growth Factor Beta; TNF-α, Tumour Necrosis Factor Alpha. When two meta-analyses reported on the same risk factor, we included the one with the largest number of studies, *, according to Lewis-Murray Obstetric Complications Scale. Edited from (Belbasis et al. 2018; Radua et al. 2018; Davies, Segre et al. 2020; Oliver, Reilly, et al. 2020; Arango et al. 2021; Gao et al. 2022; Solmi, De Toffol, et al. 2023).

2.1. Risk factors for psychosis in the general population and in individuals at CHR-P

For risk factors for psychosis, we considered a meta-umbrella review published before September 2022 that systematically reviewed the literature on risk factors for mental disorders (Arango et al. 2021), and updated it, also searching for meta-analyses reporting on risk or protective factors for psychosis in any population, including also those meeting CHR-P criteria, and grading the credibility into convincing, highly suggestive, suggestive, weak, or no significant association based on quantitative criteria accounting for several types of bias, as already extensively applied in previous works (Bortolato et al. 2017; Dragioti et al. 2019; Kim et al. 2019; Solmi, Veronese, et al. 2020; Arango et al. 2021; Solmi, Civardi, et al. 2021; Solmi, Dragioti, Arango, et al. 2021; Solmi, Dragioti, Croatto, Radua, Borgwardt, Carvalho, Demurtas, Mosina, Kurotschka, Thompson, et al. 2021; Solmi, Radua, et al. 2021). The criteria to grade the credibility of the evidence are detailed in the eMethods section. Modifiable factors with convincing evidence were considered as potential outcomes in the recommendation section.

2.2. Definition of CHR-P, and assessment tools

We described the structure of the two most frequently used tools to measure CHR-P status, namely the Comprehensive Assessment of At-Risk Mental State (CAARMS) (Yung et al. 2005), and Structured Interview for Prodromal Symptoms (SIPS) (Miller et al. 2003), their subscales and symptom domains, the criteria of each tool for different CHR-P subgroups, and differences with the brief psychotic disorder listed in the Diagnostic and Statistical Manual of Mental Disorders

5-Text Revision (DSM-5-TR) (APA 2022). In addition, given the emerging international partnerships and ongoing harmonisation efforts in the field, we conducted a systematic review to identify articles on international statements and consortia introducing novel psychometric tools for measuring CHR-P status. The tools to assess CHR-P based on basic symptoms were also mentioned given their longstanding historical relevance and their potential role in future research. Regarding the prognostic accuracy of psychometric instruments to assess the risk of psychosis, we identified a systematic review with meta-analysis published in September 2022.

2.3. Biological correlates of clinical high risk of psychosis

We conducted a meta-umbrella review reporting on the biological factors associated with CHR-P status. In addition, we conducted an umbrella review of previous meta-analyses and large consortium brain imaging studies to identify overall and regional structural, functional, metabolic, and electric differences between the brain of subjects with CHR-P and healthy controls.

2.4. Comorbid mental disorders of clinical high risk of psychosis

A subgroup among authors of these guidelines conducted a systematic review of the prevalence of comorbid mental disorders in CHR-P that has been published after September 2022 (Solmi, Soardo, et al. 2023) and included over 100 studies. Hence, no additional evidence synthesis was conducted for this subject. For the scope of these guidelines, we report the pooled prevalence of concomitant mental disorders in

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

→ 7

subjects meeting criteria for CHR-P status, to inform the assessment and care to be offered in services for prevention of psychosis. We also descriptively report the impact of concomitant mental disorders on functioning in CHR-P status, on transition to psychosis after meeting CHR-P criteria. Finally, we describe the course over time of the prevalence of comorbid mental disorders after meeting CHR-P criteria at baseline.

2.5. Age at onset of psychosis, epidemiology of transition from clinical high risk of psychosis to psychosis, psychosis prediction models, and ethical considerations in estimating and communicating risk to inform clinical implementation of psychosis prevention

Meta-epidemiological evidence on the age at onset of psychosis and transition rates from CHR-P to psychosis should inform the age threshold to access psychosis prevention services, clinical decision-making, and resource allocation. A subgroup among authors of these guidelines recently conducted large scale (i.e. more than 100 epidemiologic studies included) meta-analyses on age at onset of mental disorders (Solmi, Radua, et al. 2022), and on the risk of transition to psychosis in subjects with CHR-P status over time (Salazar de Pablo, Davies, et al. 2021; Salazar de Pablo, Radua, et al. 2021). Hence, no additional evidence synthesis was conducted for this subject. Also, we updated a systematic review reporting on prediction and prognosis models with a proper internal or external validation in mental health research (Salazar de Pablo, Studerus, et al. 2021), narrowing our focus on CHR-P status and prediction of psychosis onset. Finally, we included the published report of a previous workshop specifically focusing on ethical considerations around risk estimates and communicating risk of transition to psychosis (Fusar-Poli, Manchia, et al. 2022).

2.6. Randomised controlled trials to prevent psychosis

Given that umbrella reviews miss including individual RCTs that have been published after the most updated systematic reviews (Mei et al. 2021; Zheng et al. 2022), we also conducted a PRISMA 2020 (Page et al. 2021)-compliant systematic review of RCTs for prevention of psychosis. Recent RCTs might change the conclusions based on previous (network) meta-analyses. Given the rapid expansion of this research field, our

systematic review also included ongoing RCTs and those without available results, highlighting evolving trends. We searched Embase, PsyclInfo, MEDLINE, CINAHL, and clinical trials.gov up to September 19th, 2024. The search key is reported in eTable 1. Inclusion criteria were RCTs including subjects with CHR-P status according to explicit criteria. No restrictions were applied to the specific instrument used to define CHR-P, the type of intervention or control condition, and outcome of interest. We extracted author-defined primary and secondary outcomes. The primary outcome of this systematic review was transition to psychosis, in order to inform the guidelines for the indicated prevention of psychosis.

2.7. Meta-analyses and network meta-analyses of randomised controlled trials to prevent psychosis

We updated a previous umbrella review of (network) meta-analyses that included RCTs in subjects with CHR-P (Fusar-Poli, Davies, et al. 2019) conducted by a subgroup of the authors, following the PRIOR guidance (Gates et al. 2022). Inclusion criteria were systematic reviews with (network) meta-analyses of RCTs including subjects with CHR-P status according to explicit criteria. No restrictions were applied to the specific instrument used to define CHR-P, the type of intervention or control condition, and outcome of interest. We extracted author-defined primary and secondary outcomes. The primary outcome of this umbrella review was transition to psychosis, in order to inform the guidelines for the indicated prevention of psychosis.

2.8. Studies comparing psychosis prevention services with other services, and cost-effectiveness of psychosis prevention services

We updated a systematic review of cohort studies comparing clinical outcomes of psychosis prevention services versus other services (Sizer et al. 2022).

We conducted a PRISMA 2020 (Page et al. 2021) compliant systematic review on cost-effectiveness of psychosis prevention services. Inclusion criteria were cohort studies or simulated cohort studies focusing on cost-effectiveness of psychosis prevention services. No restrictions were applied to the specific instrument used to define CHR-P status, the type of control service, outcome of interest, and cost-effectiveness analysis. We extracted author-defined primary and secondary outcomes.

8 M. SOLMI ET AL.

2.9. Clinical recommendations development

The recommendations were developed based on the available evidence and circulated to the panel for voting on a 5-point Likert scale (1=completely disagree to 5=completely agree). Consensus was defined as at least 80% of panel members rating 4 or 5. Feedback on the whole guidelines was also collected via email. Recommendations not reaching consensus were planned to be revised and resubmitted for voting until consensus was achieved.

3. Results

3.1. Literature review

3.1.1. Risk factors for psychosis in the general population and in individuals at CHR-P

Regarding risk factors for schizophrenia, Table 2 reports on their effect size, and the credibility of evidence (Table 2). A meta-umbrella review (Arango et al. 2021) pooled evidence from umbrella reviews mapping risk factors of all mental disorders, including schizophrenia (Belbasis et al. 2018; Radua et al. 2018). We updated that meta-umbrella review. The study selection flow is reported in eFigure 1. Studies excluded after full-text assessment, with reason for exclusion, are reported in eTable 2. We identified recent umbrella reviews on psychological trauma, unsaturated fatty acids and cannabis use association with a range of mental disorders, including schizophrenia (Gao et al. 2022; Hogg et al. 2023; Solmi, De Toffol, et al. 2023), as well as a meta-analysis on prenatal and perinatal factors associated with schizophrenia published after the two former (meta-) umbrella reviews (Davies, Segre, et al. 2020; Gao et al. 2022; Solmi, De Toffol, et al. 2023), as well as one on risk factors within those meeting CHR-P criteria (Oliver, Reilly, et al. 2020). Across meta-analyses, several factors have been nominally significantly associated with psychosis, yet most meta-analytic evidence did not meet convincing or highly suggestive credibility according to established quantitative criteria (loannidis 2009; Fusar-Poli and Radua 2018; Solmi, Dragoti, Croatto, Radua, Borgwardt, Carvalho, Demurtas, Mosina, Kurotschka, Shin, et al. 2021). In more detail, as shown in Table 2, 146 factors were nominally associated with risk of psychosis, and specifically 37 were individual medical, psychiatric, psychological factors, 32 biomarkers, 16 related with pregnancy, 12 maternal and paternal factors, 10 ethnicity or migration-related factors, nine dealt with foetal growth and gestational age, six with labour season, labour complications, and delivery complications, six

with individual infections, eight with individual or environmental stress and finally 10 risk factors were identified within subjects at CHR-P. Overall, only eight factors were supported by convincing evidence, namely UHR (OR = 9.32, 95%CI 4.91–17.32), black-Caribbean ethnicity in England (OR = 4.87, 95%CI 3.96–6.00), famine during pregnancy (OR = 1.61, 95%CI 1.51–1.71), birth weight below 2kg (OR = 1.84, 95%CI 1.53–2.22), 2.5 kg (OR-1.53, 95%CI 1.31–1.78), 3kg (OR = 1.23, 95%CI 1.15–1.31), and small for gestational age at birth (OR = 1.40, 95%CI 1.25–1.57), plus cannabis use in sensitivity analyses restricted to prospective studies. Additionally, in population-based studies, 14 factors met highly suggestive criteria, including childhood adversities in sensitivity analyses restricted to prospective studies (Arango et al. 2021; Solmi, De Toffol, et al. 2023). Finally, within CHR-P subjects no specific risk factor for transition to psychosis met convincing criteria.

Importantly, only three risk factors for schizophrenia-spectrum disorders had transdiagnostic salience across other mental disorders, according to TRANSD (Fusar-Poli, Solmi, et al. 2019) criteria. These are childhood adversities (shared with borderline personality disorder, and bipolar disorder), any psychological trauma (shared with bipolar disorder and anxiety disorder) and tobacco smoking (shared with opioid use disorder) (Arango et al. 2021), suggesting psychosis has some specific environmental risk factors compared with other disorders with only partial overlap.

3.1.2. Definition of CHR-P, and assessment tools

The UHR concept, now more frequently defined as CHR-P, was introduced more than 25 years ago in Melbourne, Australia (Yung et al. 2005) (Table 3). Whilst the concept of 'prodrome' has prompted researchers to look for an 'at-risk' status with symptoms before the onset of schizophrenia, the concept of 'risk' status is different from 'prodromal' status. Whilst 'prodrome' is a retrospective and deterministic construct (i.e. the period before schizophrenia onset can only later be recognised as a prodrome). In contrast, 'risk' is a probabilistic one (i.e. those at risk have a certain probability of developing psychosis over time), that inherently implies the possibility of prevention. Hence, the concept of 'risk' also calls for caution in communicating and treating CHR-P status, as a substantial proportion – indeed, the majority – of individuals 'at risk' will not develop psychosis (Salazar de Pablo, Radua, et al. 2021).

The most widely accepted interviews to identify CHR-P are the CAARMS (Yung et al. 2005), and SIPS (Miller et al. 2003), that identify three CHR-P groups, namely Brief and Limited Intermittent Psychotic

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

9

Table 3. Main characteristics of comprehensive assessment of At-risk mental state and structured interview for prodomal symptoms interviews.

||SIPS 5.6.1 (cOPS, pOPS)|CAARMS 2015| |---|---|---| |Subscales/symptoms domains|P1: Unusual Thought Content/Delusional IdeasP2: Suspiciousness/Persecutory IdeasP3: Grandiose IdeasP4: Perceptual Abnormalities/HallucinationsP5: Disorganised Communication|P1: Unusual Thought ContentP2: Non-Bizarre IdeasP3: Perceptual AbnormalitiesP4: Disorganised Speech| |Severity|Seven levels (0–6)|Seven levels (0–6)| |Frequency|Three levels (1–3)|Seven levels (0–6)| |APS||| |Severity, frequency, functioning|Severity 3–5 plus Frequency 2|Subthreshold intensitySeverity 3–5 (P1/2), 3–4 (P3), 4–5 (P4) plus Frequency 3–6ORSubthreshold frequencySeverity 6 (P1/2/4), 5–6 (P3) plus Frequency 3ANDBoth with SOFAS drop ≥ 30% sustained ≥1 month within thepast year/<50 last 12monthsLast 12months| |Symptoms to rate|Past month, in last 12months|Up to five years earlier| |Symptoms started|Started or worsened last 12months, but could havebeen there even earlier|| |BLIPS||| |Severity, frequency, functioning|Severity 6 plus Frequency 1|Severity 6 (P1/2/4), 5–6 (P3) plus Frequency 4–6ANDSOFAS drop ≥ 30% sustained ≥1 month within the pastyear/<50 last 12monthsLast 12months| |Symptoms to rate|Past three months|Up to five years earlier| |Symptoms started|Past three months|Not more than seven days| |Duration|Not more than three months|| |GRD||| |Criteria and functioning|Schizotypal personality disorderORFirst-degree relative with psychotic disorderANDGAF drop ≥30% over the past month relative to12months prior|Schizotypal personality disorderORFirst-degree relative with psychotic disorderANDSOFAS drop ≥ 30% sustained ≥1 month within the pastyear/<50 last 12months| |Psychosis||| |Psychosis criteria|Severity 6 plus Frequency 3ORSymptoms are seriously disorganising and dangerousNA|Severity 6 (P1/2/4), 5–6 (P3) plus Frequency 4–6| |Duration||More than seven days| |Exclusion criteria||| |Exclusion criteria|Previous diagnosis of psychosis, symptoms are seriouslydisorganising and dangerousSymptoms are strongly connected with substance abuse|Previous diagnosis of psychosis (CAARMS or other sources)Symptoms strongly connected with hallucinogens,amphetaminees, and cocaine (or other known to inducepsychosis, excluding cannabinoids)| ||Symptoms better explained by other disorder||

Legend. SPS: Attenuated Psychotic Symptoms; BLIPS: Brief Limited Intermittent Psychotic Symptoms; CAARMS: Comprehensive Assessment of At-Risk Mental State; COPS: Criteria of Prodromal Syndromes; GRD: Genetic Risk and Deterioration; POPS: Presence of Psychotic Symptoms; SIPS: Structured Interview for Prodromal Symptoms. Edited from (Miller et al. 2003; Yung et al. 2005; Addington, Woods, et al. 2023).

Symptoms (BLIPS), Attenuated Positive Symptoms (APS), Genetic Risk and Deterioration (GRD) syndrome (Table 3). Briefly, BLIPS is defined by full positive psychotic symptoms that are, however, limited in frequency and duration. APS is defined by severity-subthreshold positive symptoms. GRD have schizotypal personality disorder or a first degree relative with psychosis as well as a recent drop in psychosocial functioning. Both CAARMS and SIPS should be administered to help-seeking subjects, and information gathered from the interview must be integrated with informants' input. Differences exist between CAARMS and SIPS on the maximum duration of subthreshold psychotic symptoms before defining a full-blown psychotic episode (Miller et al. 2003). Also, a

non-negligible portion of CAARMS-defined BLIPS meets full psychosis criteria for SIPS (Miller et al. 2003; Fusar-Poli, Cappucciati, et al. 2016). Moreover, the two interviews assess differently concomitant substance use as well as functional impairment for APS and BLIPS. Despite these and other differences (Table 3), CAARMS and SIPS have an overall agreement of 86-93% (Fusar-Poli, Cappucciati, et al. 2016) in determining CHR-P status, and the CONVERT tool (Fusar-Poli, Cappucciati, et al. 2016) is freely available to convert scores between the two instruments.

The difference between CHR-P (at-risk status) and brief psychotic disorder (full-blown schizophrenia-spectrum disorder) according to the DSM-5-TR (APA 2022) is

10 M.SOLMI ET AL.

blurred, as the latter includes positive psychotic symptoms lasting from one day to one month (Miller et al. 2003), which overlaps with a subgroup of patients meeting CHR-P BLIPS criteria. Hence, differentiating between CHR-P and brief psychotic disorder is not always possible and at best challenging without using a semi-structured interview.

The study selection flow of the systematic review of novel psychometric tools for measuring CHR-P status is reported in eFigure 2. The systematic review identified four studies (Brady and Larrauri 2023; Addington, Woods, et al. 2024; Wannan et al. 2024; Woods SW et al. 2024) reporting on a newly developed instrument, the Positive Symptoms for CAARMS Harmonised with SIPS (PSYCHS), which integrates CAARMS and SIPS (eTable 3). In the PSYCHS, full harmonisation of CAARMS and SIPS has been achieved for rating positive symptoms severity and determining transition to psychosis, but not for assessing CHR-P subgroups, which still relies on separate SIPS or CAARMS criteria. The harmonisation process is described in more detail in eTable 3.

In CHR-P subjects, transition to frank psychosis can be assessed using either the CAARMS, SIPS, or PSYCHS, all of which require at least one positive psychotic symptom occurring several times a week. However, for determining transition, these measures differ in symptom duration (one week for the CAARMS and PSYCHS, one month for the SIPS) and waive duration and frequency if the symptom is imminently dangerous (PSYCHS) or either seriously disorganising or dangerous (SIPS), or in case a new or increased antipsychotic has been administered (CAARMS and PSYCHS).

In addition to the UHR/CHR-P criteria operationalised by CAARMS and SIPS, a complementary approach to identifying individuals with increased risk of psychosis has been developed based on the assessment of basic symptoms using the Schizophrenia Proneness Instrument—Adult (SPI-A) (Schultze-Lutter et al. 2012) and—Child and Youth (SPI-CY) (Schultze-Lutter et al. 2012). This approach focuses on self-experienced, subtle cognitive and perceptive disturbances, which are often subjectively distressing, preceding and accompanying attenuated or brief psychotic symptoms. However, meta-analytic evidence on the prognostic accuracy of such tools is lacking.

The systematic review with meta-analysis on the prognostic accuracy of a range of psychometric tools to measure CHR-P status showed that they have excellent prognostic accuracy in predicting psychosis in clinical samples at a mean follow-up of 34 months, i.e. AUC = 0.85 (95% CI: 0.81–0.88). More in detail, sensitivity was outstanding (0.93, 95% CI: 0.87–0.96), but

specificity was poor (0.58, 95% CI: 0.50–0.66) (Oliver et al. 2022). In the discussion of that systematic review, authors indicated that CHR-P should be assessed in help-seeking, or risk-enriched populations only, because the prevalence of psychosis in the group undergoing CHR-P assessment determines CAARMS prognostic accuracy (Fusar-Poli, Schultze-Lutter, et al. 2016; Fusar-Poli, Sullivan, et al. 2019). Hence, administering CAARMS or SIPS to individuals who are not help-seeking, or outside of clinical services (such as screening in schools), is not useful in determining any risk of psychosis.

3.1.3. Biological correlates of clinical high risk of psychosis

The study selection flow of the meta-umbrella on biological correlates in individuals at CHR-P is reported in eFigure 3 (Table 4). Studies excluded after full-text assessment, with reason for exclusion, are reported in eTable 4. The meta-umbrella review identified one umbrella review reporting on the biological alterations in subjects with CHR-P (Fusar-Poli et al. 2020) (Table 4). Biological alterations in CHR-P individuals resemble those consistently shown in individuals with schizophrenia. Compared with healthy controls, subjects at CHR-P have increased cortisol levels (Chaumette et al. 2016) as well as elevated IL-6 and IL-1beta (Park and Miller 2020).

The study selection flow of meta-analyses and large consortium brain imaging studies is reported in eFigure 4. Studies excluded after full-text assessment, with reason for exclusion, are reported in eTable 5. The umbrella review identified 17 eligible studies (Davis et al. 2014; Dutt et al. 2015; Erickson et al. 2016; Harrisberger et al. 2016; Merritt et al. 2016; Walter et al. 2016; Ding et al. 2019; Wenneberg et al. 2020; Del Fabro et al. 2021; Jalbrzikowski et al. 2021; Baldwin et al. 2022; Luna et al. 2022; Vissink et al. 2022; Zhao Y et al. 2022; Simmonite et al. 2023; Zeng et al. 2023; Allen et al. 2024) reporting on brain imaging findings in subjects with CHR-P, applying different techniques (Table 4). Among these, three studies (Jalbrzikowski et al. 2021; Baldwin et al. 2022; Allen et al. 2024) were from the ENIGMA consortium (n=2,577-3,169). These studies reported on structural measures (brain volume, cortical thickness, surface area), functional measures (functional activity), and positron emission tomography or spectroscopy metrics (combined glutamate and glutamine, glutamate). In CHR-P subjects, there is converging evidence of a neuroprogressive process (Davis et al. 2014; Chaumette et al. 2016), evidenced by reduced brain volume, cortical thickness and surface area of the whole brain and specific areas (Jalbrzikowski

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

11

Table 4. Biological correlates of clinical high risk for psychosis.

|Biological factor|Source|Effect size* vs healthy controls| |---|---|---| |Inflammation vs HC||| |Cortisol (Chaumette et al. 2016)|MA (k=8, n=1,060) (Chaumette et al. 2016)|Hedge's g=0.59 (NA)| |IL-6 (Park and Miller 2020)|MA (k=5, n=229) Park and Miller 2020|SMD = 0.31 (95%CI 0.02-0.59)| |L1-β (Park and Miller 2020)|MA (k=2, n=74) (Park and Miller 2020)|SMD = -0.66 (95%CI -1.27 to -0.05)| |Neuroimaging*||| |Structural||| |Overall||| |Mean CT (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.18 (95%CI -0.25 to -0.11)| |Estimated intracranial VOL (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.13 (95%CI -0.20 to -0.06)| |Total SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.15 (95%CI -0.22 to -0.08)| |Lower PBSI-SA (Baldwin et al. 2022)|ENIGMA (n=2,644) (Baldwin et al. 2022)|t(27842) = -5.99, p<0.01)| |Lower PBSI-CT (Baldwin et al. 2022)|ENIGMA (n=2,790) (Baldwin et al. 2022)|(t(27881) = -9.11, p<0.01)| |Lower PBSI-SV (Baldwin et al. 2022)|ENIGMA (n=2,735) (Baldwin et al. 2022)|(t(2733) = -4.34, p<0.01)| |Global CT (Allen et al. 2024)|ENIGMA (n=2577) (Allen et al. 2024)|Cohen's d=-0.22CI NA (decreased)| |Infranormal global ADS (Allen et al. 2024)|ENIGMA (n=2577) (Allen et al. 2024)|$χ^{2}$ = 12.82, FDR corrected p=6.85×$10^{-4}$| |Infranormal SV ADS (Allen et al. 2024)|ENIGMA (n=2577) (Allen et al. 2024)|$χ^{2}$ = 5.68, FDR corrected p=0.02| |Infranormal SA ADS (Allen et al. 2024)|ENIGMA (n=2577) (Allen et al. 2024)|$χ^{2}$ = 16.01, FDR corrected p=2.53×$10^{-4}$| |Specific areas||| |Left rostral anterior cingulate CT (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.11 (95%CI -0.18 to -0.04)| |Bilateral posterior cingulate CT (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.13 (95%CI -0.2 to -0.06)| |Bilateral posterior central CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.12 (95%CI -0.19 to -0.05)| |Bilateral paracentral CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.15 (95%CI -0.2 to -0.08)| |Bilateral precentural CT (Jalbrzikowski et al. 2021) (LR)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.15 (95%CI -0.2 to -0.06)| |Bilateral preconueus CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.15 (95%CI -0.22 to -0.08)| |Bilateral supramarginal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.11 (95%CI -0.18 to -0.04)| |Left parahippocampal CT (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.13 (95%CI -0.2 to -0.06)| |Bilateral insula CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.17 (95%CI -0.24 to -0.1)| |Bilateral medial orbitofrontal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.13 (95%CI -0.2 to -0.06)| |2021 (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.09 (95%CI -0.16 to -0.02)| |Right lateral orbitofrontal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.13 (95%CI -0.2 to -0.06)| |Bilateral rostral middle frontal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.12 (95%CI -0.19 to -0.04)| |Bilateral caudal middle frontal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.11 (95%CI -0.18 to -0.04)| |Bilateral superior frontal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.17 (95%CI -0.24 to -0.1)| |Bilateral fusiform CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.12 (95%CI -0.19 to -0.05)| |Bilateral superior parietal CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.16 (95%CI -0.24 to -0.09)| |Bilateral inferior parietal CT (Jalbrzikowski et al. 2021) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.15 (95%CI -0.22 to -0.08)| |Bilateral inferior temporal CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.18 (95%CI -0.26 to -0.11)| |Bilateral middle temporal CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.11 (95%CI -0.18 to -0.04)| |Right transverse temporal CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.15 (95%CI -0.22 to -0.08)| |Bilateral superior temporal CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.13 (95%CI -0.2 to -0.06)| |Bilateral banks of the superior temporal sulcus CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.16 (95%CI -0.23 to -0.09)| |Bilateral lateral occipital CT (Jalbrzikowski et al. 2021) (R)|ENIGMA (n=3,169) (Zhao Y et al. 2022)|(z=-1.01, p<0.001)| |Bilateral lateral prefrontal cortex CT (Zhao Y et al. 2022) (L)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.11 (95%CI 0.04 to 0.18)| |Hippocampus VOL (Harrisberger et al. 2016)|MA (k=3, n=155) (Harrisberger et al. 2016)|Hedge's g=-0.38 (95%CI -0.73 to -0.03)| |Hippocampus VOL (R) (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrizowski et al. 2021)|Cohen's d=-0.16 (95%CI -0.23 to -0.08)| |Hippocampus VOL (R) (Walter et al. 2016)|MA (n=1,429) (Walter et al. 2016)|Hedge's g=-0.24, p=0.0418| |Hippocampus VOL (L) (Vissink et al. 2022)|MA (k=7, n=1,088) (Vissink et al. 2022)|p=0.04| |Hippocampus VOL (R) (Vissink et al. 2022)|MA (k=7, n=1,088) (Vissink et al. 2022)|Hedge's g=-0.33 (95%CI -0.60 to -0.07), p=0.01| |Thalamus VOL (Harrisberger et al. 2016)|MA (k=3, n=155) (Harrisberger et al. 2016)|Hedge's g=-0.60 (95%CI -0.96 to -0.25)**| |Thalamus VOL (R) (Ding et al. 2019)|MA (k=4, n=1,331) (Ding et al. 2019)|z=1.039, increased**| |Bilateral median cingulate VOL (Ding et al. 2019)|MA (k=14, n=1,331) (Ding et al. 2019)|z=1.034, increased| |Right fusiform gyrus VOL (Ding et al. 2019)|MA (k=4, n=1,331) (Ding et al. 2019)|z=-1.01, (increased)| |Right gyrus rectus VOL (Ding et al. 2019)|MA (k=14, n=1,331) (Ding et al. 2019)|z=-1.209, (increased)| |Medial frontal gyrus VOL (Luna et al. 2022)|MA (k=41, n=4,846) (Luna et al. 2022)|Hedge's g=-0.28, p<0.001| |Right superior frontal gyrus VOL (Ding et al. 2019)|MA (k=14, n=1,331) (Ding et al. 2019)|z=-2.321, (decreased)| |Left superior frontal gyrus VOL (Ding et al. 2019)|MA (k=14, n=1,331) (Ding et al. 2019)|z=-2.228, (decreased)| |Left superior temporal gyrus VOL (Ding et al. 2019)|MA (k=14, n=1,331) (Ding et al. 2019)|z=-1.048, (increased)| |Lateral Ventricles VOL (Vissink et al. 2022)|MA (k=3, n=841) (Vissink et al. 2022)|Hedge's g=-0.17 (95%CI 0.02 to 0.31), p=0.02|

(Continued)

12

M. SOLMI ET AL.

Table 4. Continued.

|Biological factor|Source|Effect size* vs healthy controls| |---|---|---| |Cerebrospinal VOL (Vissink et al. 2022)|MA (k=4, n=922) (Vissink et al. 2022)|Hedge's g=0.21 (95% CI 0.08 to 0.35), p=0.002| |Whole brain VOL (Vissink et al. 2022)|MA (k=7, n=1,237) (Vissink et al. 2022)|Hedge's g=-0.14 (95% CI -0.26 to -0.03), p=0.02| |Left entorhinal SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.09 (95%CI -0.16 to -0.02)| |Right isthmus cingulate SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=0.09 (95%CI 0.2 to 0.16)| |Left pars opercularis SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.09 (95%CI -0.16 to -0.02)| |Right pericalcarine SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.09 (95%CI -0.17 to -0.02)| |Left supramarginal SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.11 (95%CI -0.18 to -0.04)| |Left rostral anterior cingulate SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.12 (95%CI -0.19 to -0.04)| |Right superior frontal SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.09 (95%CI -0.16 to -0.02)| |Right inferior parietal SA (Jalbrzikowski et al. 2021)|ENIGMA (n=3,169) (Jalbrzikowski et al. 2021)|Cohen's d=-0.10 (95%CI -0.17 to -0.03)| |Functional||| |Salience network connectivity (Del Fabro et al. 2021)|MA (k=16, n=1,484) (Del Fabro et al. 2021)|Hedge's g=-0.085, p=0.012| |||FDR corrected= 0.07| |Right precuneus (Luna et al. 2022)|MA (k=55, n=3,096) (Luna et al. 2022)|Hedge's g=-0.4, p<0.001| |Bilateral anterior cingulate cortex (Luna et al. 2022)|MA (k=55, n=3,096) (Luna et al. 2022)|Hedge's g=-0.33, p<0.001| |Right superior frontal gyrus (Dutt et al. 2015)|MA (k=11, n=NA) (Dutt et al. 2015)|p<0.01 (increased)| |Right inferior frontal gyrus (Luna et al. 2022)|MA (k=55, n=3,096) (Luna et al. 2022)|Hedge's g=-0.34, p<0.001| |Left superior frontal gyrus (Luna et al. 2022)|MA (k=55, n=3,096) (Luna et al. 2022)|Hedge's g=-0.33, p<0.001| |Left medial frontal gyrus (Dutt et al. 2015)|MA (k=15, n=NA) (Dutt et al. 2015)|p<0.01 (increased)| |Right inferior parietal lobule (Dutt et al. 2015)|MA (k=15, n=NA) (Dutt et al. 2015)|p<0.01 (decreased)| |Left superior temporal gyrus (Dutt et al. 2015)|MA (k=11, n=NA) (Dutt et al. 2015)|p<0.01 (increased)| |Medial prefrontal cortex and anterior cingulate cortex activation in response to monetary stimuli (Zeng et al. 2023)|MA (k=13, N=744) (Zeng et al. 2023)|SMD=-1.330, p<0.001 (increased)| |Left cerebellum and parahippocampal gyrus activation in response to monetary stimuli (Zeng et al. 2023)|MA (k=13, N=744) (Zeng et al. 2023)|SMD=-2.051, p<0.001 (decreased)| |Right supramarginal gyrus activation in response to monetary stimuli (Zeng et al. 2023)|MA (k=13, N=744) (Zeng et al. 2023)|SMD=-1.775, p<0.001 (decreased)| |Right insula and putamen activation in response to monetary stimuli (Zeng et al. 2023)|MA (k=13, N=744) (Zeng et al. 2023)|SMD=-1.485, p<0.001 (decreased)| |Mismatch negativity||| |Mismatch negativity (Erickson et al. 2016)|MA (k=16, N=NA) (Erickson et al. 2016)|Cohen's d=0.40 (95%CI 0.23 to 0.58) (increased)| |PET/MRS||| |Medial frontal Glx (MRS) (Merritt et al. 2016)|MA (k=8, N=386) (Merritt et al. 2016)|Hedge's g=0.26 (95%CI 0.05 to 0.46)| |Frontal Glx (MRS) (Wenneberg et al. 2020)|MA (k=4, N=140) (Wenneberg et al. 2020)|SMD= 0.55 (95% CI 0.21 to 0.89), p=0.001 (increased)| |Thalamus glutamate (MRS) (Wenneberg et al. 2020)|MA (k=3, N=218) (Wenneberg et al. 2020)|SMD=-0.5 (95% CI -0.78 to -0.23, p<0.001|

Legend. ADS: average deviation score; CT: cortical thickness; Glx: combined glutamate and glutamine; IL: interleukin; MA: meta-analysis; MRS: proton magnetic resonance spectroscopy; PBS: person-based similarity index; PET: positron emission tomography; SMD: standardised mean difference; VR: Variability ratio; *Positive values indicate higher levels in CHR,** opposite findings from two works on thalamus; # when bilateral, left or right side with largest difference is indicated [L] or [R]), and reported. Edited from (Davis et al. 2014; Dutt et al. 2015; Chaumette et al. 2016; Erickson et al. 2016; Harrisberger et al. 2016; Merritt et al. 2016; Walter et al. 2016; Ding et al. 2019; Fusar-Poli Paolo et al. 2020; Park and Miller 2020; Wenneberg et al. 2020; Del Fabro et al. 2021; Jalbrzikowski et al. 2021; Baldwin et al. 2022; Luna et al. 2022; Vissink et al. 2022; Zhao Y et al. 2022; Zeng et al. 2023; Allen et al. 2024).

et al. 2021); however, the effect sizes ranges from very small to small (Cohen's $d=0.13-0.22$) (Allen et al. 2024). Among specific brain regions, alterations emerged for some areas that were confirmed across more than one metric, that were replicated in more than one study on the same metric, or that had a medium effects size. Compared with healthy controls, CHR-P subjects exhibit prominently altered frontal lobes; reduced volume in bilateral superior frontal gyri (Ding et al. 2019), reduced cortical thickness (Jalbrzikowski et al. 2021), and reduced surface area (Jalbrzikowski et al. 2021), asymmetric altered activation (right increased, left decreased) (Dutt et al. 2015; Luna et al. 2022), and increased combined glutamate plus glutamine levels (Merritt et al. 2016; Wenneberg et al. 2020). Bilateral

medial prefrontal cortex has reduced cortical thickness (Zhao Y et al. 2022) and increased motivational-related activation during reward anticipation (Zeng et al. 2023). Also, the right inferior parietal lobule shows decreased cortical thickness (Jalbrzikowski et al. 2021), surface area (Jalbrzikowski et al. 2021) and function (Dutt et al. 2015). In addition, the left superior temporal gyrus shows increased volume (Ding et al. 2019), reduced cortical thickness (Jalbrzikowski et al. 2021), and increased activation (Dutt et al. 2015). Furthermore, the left anterior cingulate has reduced cortical thickness (Jalbrzikowski et al. 2021), reduced surface area (Jalbrzikowski et al. 2021), and reduced activation in resting state (Luna et al. 2022), whereas there is increased activation in the bilateral anterior cingulate

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

13

cortex during reward anticipation (Zeng et al. 2023). Among subcortical structures, a number of studies reported a reduced volume of both the right hippocampus (Harrisberger et al. 2016; Walter et al. 2016; Jalbrzikowski et al. 2021) and the left hippocampus (Harrisberger et al. 2016; Vissink et al. 2022). The largest difference has been reported for the thalamus, which has reduced volume in CHR-P subjects (medium effect size) (Harrisberger et al. 2016), as well as reduced glutamate levels (Wenneberg et al. 2020). Lastly, bilateral lateral ventricles showed increased volume in two studies (Jalbrzikowski et al. 2021; Vissink et al. 2022). Mismatch negativity was increased in CHR-P compared with healthy controls (Erickson et al. 2016). Apart from these areas where brain alterations were confirmed across different reports or measures, all other differences were largely non-significant, not replicated, and virtually all of them had a very small effect size (Cohen's d/Hedge's g<0.2).

3.1.4. Comorbid mental disorders of CHR-P

Subjects meeting CHR-P criteria frequently have poor functioning, or a drop in functioning (at least according to CAARMS), and present with frequent symptoms of or full-blown comorbid mental disorders (Table 5). In a recent meta-analysis (Solmi, Soardo, et al. 2023), that pooled data from 312 observational and interventional studies that reported comorbid mental disorders in almost 35,000 CHR-P subjects, 75.7% of subjects met criteria for at least one non-psychotic disorder (Table 5). The most frequent disorders were

Table 5. Course and impact of concomitant mental disorders in subjects at CHR-P.

|Course of concomitant mental disorders over time in CHR-P subjects| |---| |Over 75% of CHR-P subjects have a concomitant DSM/ICD mental disorder| |Most frequent concomitant mental disorders at baseline CHR-P assessment are:| |Any mood disorder, around 43%| |Depressive disorder, around 39%| |Anxiety disorders, around 33%| |Any trauma-related disorder, around 29%| |Personality disorders, around 23%| |Up to eight years after meeting CHR-P status criteria, a comorbid mental disorder is still present in more than 50% of subjects| |Impact of concomitant mental disorders on functioning in CHR-P subjects| |Somatoform disorder, alcohol use disorder, and schizotypal personality disorder are associated with worse functioning impairment (in descending order of magnitude)| |Dysthymia and generalised anxiety disorder moderate a less severe impairment of functioning (in descending order of magnitude)| |Impact of comorbid mental disorders on transition to psychosis in CHR-P subjects| |Comorbid mood disorder, generalised anxiety disorder and agoraphobia lower the odds of transition to psychosis.| |Legend. DSM, Diagnostic and Statistical Manual; ICD, International Classification of Diseases. Edited from (Solmi, Soardo, et al. 2023).|

anxiety/mood disorders (60.1%), any mood disorders (43.7%), any depressive disorder/episode (39.3%), any anxiety disorder (33.7%), major depressive disorder (31.1%), any trauma-related disorder (29.4%), any personality disorder (23.3%), any pervasive developmental disorder (around 14.8%), neurotic, stress-related and somatoform disorders (around 14.2%), cannabis use disorder (13%), schizotypal personality disorder (12.5%), attention-deficit/hyperactivity disorder (12.5%), any behavioural disorder (12.5%), pervasive developmental disorder not otherwise specified (12.1%), social anxiety disorder (11.8%), any substance use disorder (11.8%), phobias (11.5%), avoidant personality disorder (10.5%), borderline personality disorder (10.4%), and other disorder in less than 10% of study participants. While the vast majority of these comorbid disorders do not seem to be associated with functioning (i.e. aggravate or mitigate functioning impairment already present with CHR-P status), some of them are (Solmi, Soardo, et al. 2023). Somatoform disorder, alcohol use disorder, and schizotypal personality disorder (in descending order of magnitude) are associated with a more severe impairment of functioning, whilst dysthymia and generalised anxiety disorder (again in descending order of magnitude) with a less severe impairment of functioning. Also, importantly, according to the same meta-analysis (Solmi, Soardo, et al. 2023), comorbid mood disorder, generalised anxiety disorder and agoraphobia are associated with lower odds of transition to psychosis. After meeting CHR-P criteria, up to eight years of follow-up, rates of any non-psychotic disorders decrease somewhat, yet remain high, being still present in >50% of subjects meeting CHR-P criteria at baseline.

Given the high proportion of CHR-P subjects with comorbid mental disorders, the clinical needs of these subjects go beyond the prevention of psychosis.

3.1.5. Age at onset of psychosis, epidemiology of transition from CHR-P to psychosis, psychosis prediction models, and ethical considerations to inform clinical implementation of psychosis prevention

Regarding age at onset of psychosis, according to the largest meta-analysis conducted so far on the age at onset of mental disorders (Solmi, Radua, et al. 2022), the proportion of individuals with onset of a schizophrenia spectrum disorder before the age of 14, 18, 25 was 3%, 12.3%, 47.85%, supporting that the age eligibility criteria for psychosis prevention programs should be age 14 (Table 6).

14

M. SOLMI ET AL.

Table 6. Epidemiology of transition to psychosis, risk estimates, and ethical considerations.

|Epidemiology of transition to psychosis|| |---|---| |Subjects at CHR-P have increased risk of transition to psychosis|The prevalence of schizophrenia is around 1% in the general population. The prevalence of CHR-P is 1.7% and 19.2% in the general population and in clinical samples, respectively.The rate of transition to psychosis in CHR-P subjects is as high as 8% at 6 months, reaching 29% at 4 years, then plateauing and reaching 35% at 10 years from CHR-P baseline assessment, but lower for children and adolescents| |Estimating risk of transition to psychosis|| |Tools|The two most employed diagnostic tools to assess early psychosis are CAARMS and SIPS, with good prognostic performanceIn clinical centres where CAARMS or SIPS/SOPS are not used, an individualised transdiagnostic risk calculator has been recently developed employing electronic health information. This tool can be used not only in CHR-P patients, but also in any other individual with non-psychotic psychiatric disorder.| |Population|Assessment of CHR-P status or risk of psychosis should only be assessed in help-seeking non-psychotic (sub) clinical population, as the prognostic accuracy of CAARMS and SIPS/SOPS is poor if applied to the general population.| |Validate prognostic/prediction models|Models should be developed, validated internally, and replicated in external samples to ensure generalisability. Models&#x27; prognostic/prediction accuracy must be determined through discrimination values (Harrel C index or Area-Under-the Curve, sensitivity, specificity), calibration (regression slope of prognostic index), overall performance (R$^{2}$, Brier Score), in longitudinal cohort studies.| |Risk calculators candidates|Multiple risk calculators have been developed, which should be updated with evidence on additional risk factors (see Table 2).| |Key ethical principles|| |Autonomy|The right for an individual to make their own choice| |Beneficence|The principle of acting with the best interest of the other in mind| |Non-maleficence|The principle that &#x27;above all, do no harm&#x27;| |Justice|Fairness and equality among individuals| |Explainability/interpretability|Why a clinical prediction model arrives at an output/prediction| |Implementation of key ethical principles, and reasons|| |Keep a human component in the risk assessment and communication|Promote social interactions, promoting human-to-human interaction in the risk estimate, avoiding social isolation involve the user in the critical interpretation of risk estimate, to avoid acritical acceptance of automated outputs.Allow users to benefit from risk estimates, accepting and preventing the risk of automation distrust deriving from the absence of human-to-human interaction| |Ensure validity and representativeness|Always ensure accountability, which is complicated in fully automated risk estimatesMitigate digital divide, i.e. gap in access to preventive care for those subjects with limited digital literacy or asked access to digital devicesOnly use validated clinical prediction models (see above), to avoid structural bias and over-optimismEnsure that clinical prediction models have been validated in populations representative of users (under- or over-representation of minorities limits generalisability)| |Ensure transparency and reproducibility|Use glass-box clinical prediction models with transparent and reproducible algorithms, instead of black-box &#x27;obscure&#x27; machine learning protocols| |Training of operators|Offer CHR-P assessment tools training to operatorsOffer clinical prediction model training to operators, so all variables in the clinical prediction model are accurately collected, beyond other sources of incentives that could bias diagnostic codesTrain in ethical communication of risk|

Legend. Edited from (Fusar-Poli, Rutigliano, Stahl, Davies, Bonoldi, et al. 2017; Fusar-Poli, Rutigliano, Stahl, Davies, De Micheli, et al. 2017; Fusar-Poli et al. 2018; Fusar-Poli, Werbeloff, et al. 2019; Puntis et al. 2021; Salazar de Pablo, Studerus, et al. 2021; Salazar de Pablo, Woods, et al. 2021; Fusar-Poli, Manchia, et al. 2022; Meehan et al. 2022).

A recent meta-analysis pooled data from 130 studies and 9,222 individuals (Salazar de Pablo, Radua, et al. 2021). The cumulative transition risk from CHR-P to psychosis was 8% at 6 months, 14% at 1 year, 17% at 1.5 years, 20% at 2 years, 25% at 2.5 years, 27% at 3 years, 28% at 3.5 years, 29% at 4 years, and 35% at 10 years (Salazar de Pablo, Radua, et al. 2021) (Table 6). Moreover, a lower proportion of female individuals ($\beta = -0.02$; 95% CI: -0.04 to -0.01) and a higher proportion of BLIPS ($\beta=0.02$; 95% CI: 0.01-0.03) were associated with higher transition risk. Also, across CHR-P cohorts, higher age has been identified as a risk factor for transition to psychosis (Fusar-Poli et al. 2012). Indeed, risk for transition to psychosis is lower in children and adolescents fulfilling CHR-P criteria, being 9.5% at 1 year, 12.1% at 2 years and only 16.1% at

≥5 years (Lång et al. 2022). The lower conversion rates to psychosis compared to adults are likely due to the fact that attenuated positive symptoms are less specific before adulthood, possibly being associated with internalising, externalising, substance use and personality disorder traits or diagnoses that might also emerge around the same age (Kelleher, Murtagh, et al. 2012; Gerstenberg et al. 2015; Schimmelmann et al. 2015; Gerstenberg et al. 2016). Additionally, even full psychotic symptoms appear to be less specific in youth. The meta-analytically determined median prevalence of psychotic symptoms – predominantly auditory hallucinations – is as high as 17% among children aged 9–12 years and much lower, i.e. 7.5%, among adolescents aged 13–18 years (Kelleher, Connor, et al. 2012) where epidemiologically the risk for psychosis

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

15

should be higher (Solmi, Radua, et al. 2022). Nevertheless, when measured in the same study with homogeneous definitions and tools, the duration of untreated psychosis appears to be three times as long in adolescents than adults with first episode psychosis (5.4±3.1 vs 18.7±6.2months (Stentebjerg-Olesen et al. 2016)), highlighting the need to focus on psychosis prevention in youth too. The risk of transition to psychosis can be measured with two different yet complementary approaches in subjects who are help-seeking or assessed in a clinical context. A first possibility is to administer a clinical interview to assess whether a subject meets CHR-P criteria. CAARMS's area under the curve (AUC) at 2years is 0.79 (Oliver et al. 2018), with a sensitivity and specificity of 83% and 74% at 6 months (Yung et al. 2005), respectively. SIPS had a sensitivity of 100%, as well as a specificity of 71% at six months (Miller et al. 2003) follow-up. A second approach to quantify the risk of transition to psychosis is to use risk calculators. Risk calculators for psychosis are among the most valid clinical prediction models in psychiatry, and can be used by healthcare professionals, regardless of previous administration of CAARMS or SIPS, since they rely on demographic characteristics and clinical diagnoses (Table 6). The use of calculators might be a more feasible approach in settings or time periods with insufficient staff to administer CAARMS or SIPS. The two approaches can also be combined, as the risk of transition to psychosis is not the same across all individuals with CHR-P. Clinical prediction models are crucial for developing and implementing precision psychiatry, but methodological and ethical guidance should be followed. Regarding the methodological standards for prediction models, a large sample size is needed to have a sufficient number of events (at least 100), and an appropriate events/ model variables ratio (ideally at least 20 events per model variable, i.e. event per variable, EPV), to protect from biased results and overfitting (Meehan et al. 2022). Also, the variables/predictors that compose the model should be identified via systematic review of the literature, and should be considered all together in multivariable models. Furthermore, after its development within a given sample, each model should be validated internally and externally in new cohorts of individuals (Meehan et al. 2022). The validity of each prediction model should be measured with discrimination metrics, calibration, and overall performance (Table 6) (Fusar-Poli et al. 2018). Finally, the clinical impact of each model should be measured with simulation studies or real-world experimental studies (Meehan et al. 2022). We account for these and other

methodological aspects of prediction typically measured with the Prediction model Risk Of Bias Assessment Tool (PROBAST) (Wolff et al. 2019), updating a previous systematic review focusing on prediction of psychosis. The study selection flow of the updated systematic review (Salazar de Pablo, Studerus, et al. 2021) on prediction and prognosis models with proper internal or external validation specifically focusing on CHR-P status and prediction of psychosis onset, is reported in eFigure 5. Out of 1742 hits, we ultimately identified 40 additional studies since previous reviews (Fusar-Poli, Stringer, et al. 2019; Osborne and Mittal 2019; Zarogianni et al. 2019; Zhang et al. 2019; Bourgin et al. 2020; Collin et al. 2020; Oliver, Wong, et al. 2020; Perkins et al. 2020; Raket et al. 2020; Studerus et al. 2020; Yassin et al. 2020; Dickens et al. 2021; Grent-t-Jong et al. 2021; Haining et al. 2021; Hauke et al. 2021; Irving et al. 2021; Koutsouleris et al. 2021; Mongan et al. 2021; Puntis et al. 2021; Worthington et al. 2021; Antonucci et al. 2022; Dwyer et al. 2022; Koutsouleris et al. 2022; Lee et al. 2022; Li Z et al. 2022; Moore et al. 2022; Sullivan et al. 2022; Worthington et al. 2022; Doborjeh et al. 2023; Iftimovici et al. 2023; Loch et al. 2023; Lyu et al. 2023; Mondelli et al. 2023; Argolo et al. 2024; Byrne et al. 2024; Hartmann et al. 2024; Kizilay et al. 2024; Krakowski et al. 2024; Reinen et al. 2024; Sprungli-Toffel et al. 2024; Zhu Y et al. 2024), highlighting an increased research interest in the field. Studies excluded after full-text assessment, with reason for exclusion, are reported in eTable 6. The systematic review found that all prediction models published in the field of CHR-P status and prediction of psychosis onset were at high risk of bias according to PROBAST. The largest source of bias was the analysis domain. A critical factor was that studies tended to include too many variables per number of events, or even completely neglected this aspect, not even reporting information to assess EPV (Fusar-Poli et al. 2018). Another critical factor was the inclusion of a component of the outcome (i.e. positive symptoms) among predictors. Only 11 studies externally validated models in a sample with ≥100 events (Fusar-Poli et al. 2018). Finally, only three prediction models (Fusar-Poli, Rutigliano, Stahl, Davies, Bonoldi, et al. 2017; Raket et al. 2020; Krakowski et al. 2024) assessed clinical utility, applying decision-curve analyses, aiming to measure the potential 'net benefit' over routine practice (Fusar-Poli et al. 2018). Hence, despite methodological limitations in the field of mental health, in psychiatry the field of prevention of psychosis is the one where the most progress has been made in terms of clinical prediction. Nevertheless, the

16 M. SOLMI ET AL.

potential implementation of prediction models requires ethical considerations.

Regarding the ethical considerations around risk pred-iction and communication, a recent review and consens-sus statement involving multidisciplinary stakeholders led by the European College of Neuropsychopharmacology Prevention of Severe Mental Disorders (PSMD) Cluster and funded by the European Brain Research Area (EBRA) provided a unique ethical insight on how researchers and clinicians should assess and communicate risk calcul-lated with a novel clinical prediction model (Fusar-Poli, Manchia, et al. 2022). Such an ethical framework applies well to assessing and communicating the risk of transi-tion to psychosis in subjects meeting CHR-P criteria. Research and clinical practice should always follow key ethical principles, namely autonomy, beneficence, non-maleficence, justice, and explainability/interpretabil-ity. These principles should guide the clinical implement-ation of precision psychiatry, risk assessment and communication. Briefly, some overarching clinically relev-ant principles are necessary for an ethical implementa-tion of risk calculations and precision psychiatry in the field of prevention of psychosis (and beyond). These include offering human-to-human interaction to users, only relying on validated clinical prediction models that must be transparent and reproducible and administered by appropriately trained operators. The definition of eth-ical principles, and the reasons for their clinical im-lementation are described in more detail in Table 6.

3.1.6. Randomised controlled trials to prevent psychosis

The study selection flow of the systematic review of RCTs in subjects meeting CHR-P status criteria is reported in eFigure 6 (Table 7). Out of initial 1207 hits, we ultimately included 44 RCTs reported across 70 publications (McGorry et al. 2002; Morrison et al. 2004; McGlashan et al. 2006; Nordentoft et al. 2006; Morrison et al. 2007; Ruhrmann et al. 2007; Amminger et al. 2010; Addington et al. 2011; Bechdolf et al. 2011; Yung et al. 2011; Bechdolf et al. 2012; Morrison et al. 2012; van der Gaag et al. 2012; McGorry et al. 2013; Woods SW et al. 2013; Miklowitz et al. 2014; Kantrowitz et al. 2015; Piskulic et al. 2015; Ising et al. 2016; Loewy et al. 2016; Stain et al. 2016; Albert et al. 2017; Bechdolf et al. 2017; Cadenhead K et al. 2017; Choi et al. 2017; McGorry et al. 2017; Woods S et al. 2017; Bhattacharyya et al. 2018; Mossaheb, Papageorgiou, et al. 2018; Mossaheb, Schäfer, et al. 2018; Nelson et al. 2018; Piskulic et al. 2018; TD et al. 2018; Berger et al. 2019; Bolt et al. 2019; Cadenhead KS et al. 2019; Davies, Paloyelis, et al. 2019; Davies, Rutigliano, et al. 2019;

Pozza et al. 2019; Wessels et al. 2019; Wilson et al. 2019; Alqarni et al. 2020; Amminger et al. 2020; Appiah-Kusi et al. 2020; Davies, Wilson, et al. 2020; Glenthøj et al. 2020; Kristensen et al. 2020; Mallawaarachchi et al. 2020; Schmidt A et al. 2020; Glenthøj et al. 2021; Li H et al. 2021; Myin-Germeys et al. 2021; Pozza et al. 2021; Rinne et al. 2021; Tsai et al. 2021; Friedman-Yakoobian et al. 2022; Addington, Liu, et al. 2023; Bechdolf et al. 2023; Davies et al. 2023; McGorry et al. 2023; Tang et al. 2023; Waite et al. 2023; Zhao J et al. 2023; Zhu H et al. 2023; Davies, Bossong et al. 2024; Davies, Martins et al. 2024; Livingston et al. 2024; Qurashi et al. 2024; Wasserthal et al. 2024) with publicly available results, plus 30 RCTs that are still ongoing or for which results are not publicly available. Studies excluded after full-text assessment, with reason for exclusion, are reported in eTable 7. Details of the included RCTs are reported in Table 7, while characteristics of RCTs still ongoing or for which no results are available are described in eTable 8.

Among the RCTs that reported results, eight studies focused on cognitive therapy or CBT or CBT-informed interventions (Morrison et al. 2004; 2007; Addington et al. 2011; Morrison et al. 2012; van der Gaag et al. 2012; Ising et al. 2016; Stain et al. 2016; Pozza et al. 2019, 2021; Addington, Liu, et al. 2023; Waite et al. 2023), 14 on other psychosocial interventions (Nordentoft et al. 2006; Bechdolf et al. 2012; Miklowitz et al. 2014; Piskulic et al. 2015; Loewy et al. 2016; Albert et al. 2017; Choi et al. 2017; Piskulic et al. 2018; TD et al. 2018; Wessels et al. 2019; Glenthøj et al. 2020; Kristensen et al. 2020; Glenthøj et al. 2021; Li H et al. 2021; Myin-Germeys et al. 2021; Rinne et al. 2021; Tsai et al. 2021; Friedman-Yakoobian et al. 2022; Zhao J et al. 2023), 16 on biological interventions (McGlashan et al. 2006; Ruhrmann et al. 2007; Amminger et al. 2010; Woods SW et al. 2013; Kantrowitz et al. 2015; Cadenhead K et al. 2017; Woods S et al. 2017; Bhattacharyya et al. 2018; Mossaheb, Papageorgiou, et al. 2018; Mossaheb, Schäfer, et al. 2018; Cadenhead KS et al. 2019; Davies, Paloyelis, et al. 2019; Davies, Rutigliano, et al. 2019; Wilson et al. 2019; Appiah-Kusi et al. 2020; Davies, Wilson, et al. 2020; Schmidt A et al. 2020; Davies et al. 2023; Tang et al. 2023; Zhu H et al. 2023; Davies, Bossong et al. 2024; Davies, Martins et al. 2024; Livingston et al. 2024; Qurashi et al. 2024), and six on combined (i.e. psychosocial plus biological) interventions (McGorry et al. 2002; Bechdolf et al. 2011; Yung et al. 2011; McGorry et al. 2013; Bechdolf et al. 2017; McGorry et al. 2017; Nelson et al. 2018; Berger et al. 2019; Bolt et al. 2019; Alqarni et al. 2020; Amminger et al. 2020; Mallawaarachchi et al. 2020;

|Table 7. Individual randomised controlled trials to prevent psychosis in people at clinical high risk for psychosis.||||||||||||| |---|---|---|---|---|---|---|---|---|---|---|---|---| |Author, yearRCT/R68|Centre|Refrans|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment*|Improve from baseline within group| |Cognitive therapy or cognitive behavioural therapy (additional et al., 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 2011,|||||||||||||

2018-09-17

$p_1, p_2, \dots, p_n$

Table 7. Continued.

|Author, year/RCT/RoB|Centre|Referrals|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment*|Improve from baseline within group| |---|---|---|---|---|---|---|---|---|---|---|---|---| |(van der Gaag et al. 2012; king et al. 2016)|Mental Health Centre PyQ Haaglanden, (Haigroup Academic Medical Centre and Mental Health Centre PyQ Amsterdam, Mental Health Centre Rivindunen, Leiden Mental Health Institute Friesland, Friedland, Amsterdam.|General practitioners|CAARMS Prodomal Questionnake|201 97/104|22.9 (experimental) 22.6 (control) 14 to 35|SB|CBT for ultra high risk patients with Treatment as usual|Treatment as usual|24M|Transition depression, anxiety, quality of life, functioning, beliefs about illness|CBT UHR (Transition)|CBT for ultra-high risk with Treatment as usual (depression, anxiety, distress) (anxiety, depression)| |(Waite et al. 2023) (SRCTNBS601537 Low RoB|Oxford Health NHS Foundation Trust, Berkshire Healthcare NHS Foundation Trust, England, UK|Mental health services|CAARMS|40 21/19|16.9 14 to 23|SB|Sleep/Well therapy with Treatment as usual|Treatment as usual|3M|Insomnia severity, psychotic experiences, depressive, anxiety symptoms|Sleep/Well therapy + Treatment as usual (Insomnia severity)|NR| |Other psychosocial Interventions (Bechdolf et al. 2012; Wessels et al. 2019)|Department of Psychiatry and Psychotherapy in Cologne, Bonn, Duseldorf, Munich, Germany|Primary care, mental healthy specialists, counsellers|ERiraos, IRAOS|128 63/65|26.0 NA|OL|Integrated psychological intervention (combination of CBTS/FCCRT/ EDU) (A et al. 2010)|Supportive counselling|12M|Transition treatment response|Integrated psychological intervention (transition)|NP| |NCT00204087 High RoB|Coperham Hospital Corporation and Psychiatry Hospital Aarhus, Denmark (Petersen et al. 2005)|All inpatient and outpatient mental health services in Copenhagen (Copenhagen Hospital Corporation) and Aarhus county|ICD-10 (Schizotypal disorder)|79 42/37|24.9 18 to 45/Petersen et al. 2005)|OL|Integrated treatment (assertive community treatment, individual or group social skills training, EDU in multiple family groups)|Standard treatment|24M|Transition, psychotic, negative, disorganised symptoms|Integrated treatment (Transition/negative symptoms)|NR| |(Albert et al. 2017) High RoB|Mental Health Centre Copenhagen, Hildegaardwej Denmark|6 Danish OPUS teams (S in Copenhagen, and one in Aarhus)|ICD-10 (schizotypal disorder)|83 39/44|26.6 18 to 35|SB|Integrated treatment (family treatment, and skill training)|Treatment as usual|6M|Transition, psychotic, negative, disorganised symptoms, substance use, functioning, employment status, cognitive function|None|OPUS Treatment (Negative symptoms, functioning) (anxiety, functioning)| |(Choi et al. 2017) Some concerns|Institute of living in Harford, CT, and COPE in New York State Psychiatry Institute, Columbia University, US|Mental health specialists|SIPS SOPS|62 31/31|18.35 16 to 24|DB|Processing Speed Training (table-based cognitive training with pupilometric neurofeedback)|Placebo training|2M|Cognition (processing speed), social functioning, depressive, anxiety symptoms|PST (cognition)|PST (cognition)| |(Friedman-Yakooblan et al. 2022) High RoB|NR|Social Media, online advertisements, clinicians, schools, universities|COPS SIPS|38 20/18|19.2 (Intervention) 15 to 26|SB|Cognition for Learning and for Understanding Everyday Social Situations (CLUES)|Enriched Acceptance and Commitment Therapy (EnACT)|9M|Social functioning, cognitive|CLUES (social functioning, social cognition)|CLUES (social functioning, social cognition, positive symptoms, visual memory, reaction time)|

Table 7. Continued.

|Author, year/RCT/Rob|Centre|Referrals|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment*|Improve from baseline within group| |---|---|---|---|---|---|---|---|---|---|---|---|---| |(TD et al. 2018; Glenthaj et al. 2020; Kristensen et al. 2020; Glenthaj et al. 2021)|Mental Health Centre Copenhagen, Copenhagen, Denmark|Psychiatric in- and out-patient facilities of the amendment area of Copenhagen|CAARMS|146 73/73|23.93 (Intervention) 23.90 (Control) 18 to 40|SB|Cognitive remediation with treatment as usual|Treatment as usual|6M|Cognitive functioning, clinical symptoms, social and role functioning white matter organisation|None|NR| |NCT02098408 High Rob||||||||||||| |(LI H et al. 2021; ChITR2000031741 High Rob|Shanghai Mental Health Centre, Suzhou Guangui Hospital, Psychiatric Department, San Francisco, US|NR|SIPS SOPS|80 40/40|20.51 14 to 45|SB|SMART (Specific Memory Attention Resource and Training) app|Treatment as usual|3M|Cognitive function|SMART (Attention/ vigilance)|SMART (Attention/vigilance)| |(Loewy et al. 2016) NCT00655239 Some concerns||Mental Health specialists, primary care, general population (school, internet)|SIPS|83|18.2 12 to 30|DB|Auditory training (AT) (laptop-based auditory processing-based training)|Placebo training (computer gaming (CG)|2M|Cognitive (verbal, marmac), positive, negative, disorganised symptoms, functioning|AT (cognition, disorganised symptoms)|AT (cognition, positive, negative, disorganised symptoms, functioning)| |(Miklavitz et al. 2014; Rinne et al. 2021) NCT01907282 High Rob|NAPLS-2 consortium, Atlanta, Cambridge, Los Angeles, San Diego, Chapel Hill, New Haven, New York, US, Calgary, Canada|Health care providers, social service agencies, general population (school, internet)® SOPS|SIPS SOPS|129 66/63|17.4 12 to 32|NR|Family-focused treatment for individuals at CHR|Family EDU (Enhanced care)|6M|Positive symptoms, negative symptoms, functioning, family interaction, depression|Family-focused treatment (Positive symptoms)|Family focused therapy (Positive symptoms, negative symptoms, functioning)| |(Reininghaus et al. 2019) NTR4252 High Rob|INTERACT trial - 5 regions in Belgium and the Netherlands|Treating clinicians|CAARMS|148 71/77|25 15 to 65|SB|Acceptance and Commitment Therapy in Daily Life (ACT-DL)|Standard care|12M|Psychotic distress, social functioning, symptom severity|ACT-DL (psychotic distress, negative symptoms, global functioning)|NR| |(Pskulk et al. 2015) Some concerns|At Risk for Mental Lines Research group Calgary, Canada|Health care providers, social service agencies, general population (school, internet)|SIPS|32 18/14|19.72 (Intervention) 17.5 (Control) 14 to 35|SB|Auditory processing CRT (Brain Fitness Program)|Placebo (video games)|9M|Cognition functioning, positive, negative symptoms,|None|Placebo (cognition, CRT (cognition, functioning)| |(Pskulk et al. 2018) High Rob|At Risk for Mental Fitness Research groups, Calgary, Canada|Mental health agencies, school counseling services, community general population (oureach campaigns)|SOPS|12 7/5|19.0 (Median, Intervention) 20.0 (Median, control) 14 to 25|SB|CRT+ Motivational interviewing|CRT only|2M|Attrition rates|NR|NR| |(Ttai et al. 2021) High Rob|China Medical University Hospital, Taiwan|NR|Chinese Version of Schaltypal Personality Questionnaire-Brief (CSPQ-8)Chinese Mandarin State and Trait Anxiety Form-Y (CMSTAIY)|92 46/46|21.34 20 to 35|NR|Health-awareness-strengthening lifestyle program (HASIL)|Waitlist control|0.5M|Health promotion lifestyle, nutrition, physical exercise, quality of life, anxiety|HASL (health promotion lifestyle, nutrition, quality of life, anxiety)|HASL (self-actualization aspects of health promotion lifestyle, anxiety)|

(Continued)

Table 7. Continued.

|Author, year/RCT/RoB|Centre|Referrals|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment*|Improve from baseline within group| |---|---|---|---|---|---|---|---|---|---|---|---|---| |(Zhao I et al., 2023) ChICTR2000039623 Low RoB|Beijing Anding Hospital, China|Referral from treating psycharists, Self-referral|SIPS|57 28/29|25.5 18 to 35|SB|EMDR|Waiting list|3M|PTSD symptoms,positive, anxiety, depressive symptoms, suicidal ideation|EMDR (PTSD symptoms, positive, anxiety, depressive symptoms, suicidal ideation)|EMDR (PTSD symptoms, positive, anxiety, depressive symptoms, suicidal ideation)| |Biological interventions (Amminger et al., 2010; Mosaheb, Papageogiou, Papageogiou, et al., 2018; Mosaheb, Scahale, et al., 2018) NCT00396643 Low RoB|Psychosis detection unit, Vienna General Hospital, Austria|Mental health specialists, counsellers|PANSS, clinical interview for accounting for CAARMS frequency criteria, and GRD definition|81 41/40|16.4 13 to 25|DB|ω - 3 PUFs (1.2 grams daily) + NBI (9 sessions of pharmacological (AD, BDOZ), psychological, psychological intervention including case-management with social support)|Placebo + NBI (9 sessions of pharmacological (AD, BDOZ), psychological, psychological intervention including case-management with social support)|3M|Transition positive, negative, general, depressive symptoms, functional, drop-out, lipid profile|ω - 3 PUFs (transition, positive, negative, general symptoms, and functioning)|NR| |(Cadenhead K et al., 2017; Cadenhead KS et al., 2019) NCT01429454 Low RoB|NAPLS Consortium|Mental health specialists, schools, colleges|SIPS|127|18.69 12:30|DB|ω - 3 PUFs (700 mg EPA, 480 mg DHA)|Placebo|6M|Transition cardiometabolic, dietary Omega 3 fatty acids and oxidative stress indices|None|ω - 3 PUFs (positive, functioning)| |(Davies et al. 2023; Davies C. et al., 2024) ISRCTM46322781 High RoB (Davies Cathy et al., 2024) ISRCTM46799530 Some concerns|Institute of Psychiary, Psychology and Neuroscience, King&#x27;s College, London, UK|NR|CAARMS|52 33/19|22.7 23.9/24.1 18 to 35|DB|Single oral dose of 600 mg of Cannabidol (CBD)|Placebo|0.03M (one day)|Hipocampal blood flow* Hipocampal glutamate*|Single oral dose of 600 mg of Cannabidol (CBD) hipocampal blood flow, lipocampal glutamate*|NP| |(Bhatcharyya et al., 2018; Appiah-Kusi et al., 2020; Davies, Wilson, et al., 2020) ISRCTM46322781 Low RoB (Wilson et al., 2019) Low RoB|King&#x27;s College, London IoPPN, UK|Early intervention service|CAARMS|52 16/17|22.43 NA|DB|Cannabidol (600 mg)|Placebo*|0|fMRI BOLD signal during encoding, recall and recall-inducaminaboid, cortisol, activity of the medial temporal lobe and striatum during fear processing|Cannabidol (encoding, recall)|NP| |(Davidson, Paloyilis, et al., 2019) Low RoB|Kings College London, London UK Clinical Research Facility|Early intervention service|CAARMS|33|NR 18 to 35|DB|Cannabidol (600 mg)|Placebo, Healthy Control (Not CHR, no CBD administered)|NA (cross-sectional)|Behavioural performance (5 domains: mean, monetary reward, accuracy, reaction time, false-starts, any trial responses)|CHR-CBD (Reaction time, false starts)|NP| |(Davies, Paloyilis, et al., 2019; Davies, Ruigblano, et al., 2019; Schmidt A et al., 2020) Low RoB|NR|Two early detection services: The OAS5 and Tower Hamlets Early Detection Service|CAARMS|30|NR 18 to 35|DB|Intranasal oxytocin (40IU)|Placebo|0|Global central blood flow, hipocampal cerebral blood flow, choline in ACC, cognitive and social empathy|Intranasal oxytocin (hippocampal cerebral blood flow)|NP|

(Continued)

Table 7. Continued.

|Author, year/RCT/RoB|Centre|Referrals|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment#|Improve from baseline within group| |---|---|---|---|---|---|---|---|---|---|---|---|---| |(Kantowitz et al. 2015) NCT00826202 Low RoB|Nathan Kline Institute, Yale, Zucker Hiltiskie hospital, and New York State Psychiatry Institute, US|Mental health specialists, primary care|SOPS|44 20/24|19.5 13 to 35|DB|D-serine|Placebo|4M|Negative symptoms, positive symptoms, cognition, sleep|D-serine (negative symptoms)|D-serine (negative symptoms)| |(L-livingston et al. 2024) NCT0090483 Low RoB|Outreach and Support in South London (OASIS), South London and Maudsley NHS Foundation Trust|NR|CAARMS|66 24/22|25.3 18 to 32|DB|5 mg of oral diazepam (crossover design)|Placebo (crossover design)|0.5M|Hippocampal regional, cerebral blood flow (CRBI)|Diazepam (Hippocampal regional, cerebral blood flow)|NP| |(Qurashi et al. 2024) NCT02569307 Low RoB|Abasi Shahed Hospital, Karachi Pakistan, Civil hospital Karachi, Karachi, Pakistan, Karwi e Hayat, Karachi Pakistan, Colleges and Universities, Karchi Pakistan; General Practitioners, Karachi Pakistan, Institute of Behavioural Sciences, Karachi Pakistan|Self-referrals|CAARMS|326 82/82/82/80|24 16 to 25|DB|1. Minocycline 2. omega-3 fatty acids 3. minocycline + omega-3 fatty acids|Double placebo|6M|Transition to psychosis,positive, depressive symptoms, social and role functioning|None|omega-3 fatty acids (positive, depressive symptoms) [among non-translationers]) minocycline + omega-3 fatty acids (positive, depressive symptoms] (among non-transitioners])| |(McGlashan et al. 2006) Low RoB|PRIME clinics: Yale, North Carolina/Chapel Hill, US; Calgary, Toronto, Canada.|Clinicians, general population (advertisements)|SPS|60 31/29|17.7 12 to 45|DB|Olanzapine (5-15mg)|Placebo|12M|Transition positive, negative, general depressive, manic symptoms, functioning, drop-out|None|Olanzapine (total, positive symptoms, functioning) Placebo (functioning)| |(Ruhmann et al. 2007) High RoB|Departments of Psychiatry, Cologne, Bonn, Dussedkorf, Munich, Germany|Psychologists, psychiatrists, general practitioners, school or university counselling services, self-referrals|ERlaos ERI-BAPPSS|124 65/59|25.6 18-36|OL|Needs-focused intervention + amisulpride (50-800mg/day)|Needs-focused intervention only|24M|Positive symptoms negative, depressive, basic symptoms, functioning|Needs-focused intervention + amisulpride (positive symptoms, negative symptoms, functioning)|Need-focused intervention + amisulpride (positive symptoms, negative symptoms, functioning, depressive symptoms)| |(Tang et al. 2023) ChICTR-INR-16009566 High RoB|Shanghai Mental Health Centre, China|Shanghai At-Risk for Psychosis Program (SHARP)|SPS SOPS|65 31/27|19 15 to 45|DB|TMS|Sham TMS|0.5M|Visual spatial learning [VSL], cognitive functioning, transition to psychosis at 12M|TMS (VSL) performance, transition to psychosis at 12M)|TMS (VSL) performance and precision| |(Woods SW et al. 2013) NTC0026874 NTC00291226 Low RoB|PRIME Research Clinic, Yale University, New Haven, USA|Community education efforts|COPS SOPS|8 4/4|15.3 (Intervention) 16.5 (Control) 15 to 17|DB|Glycine (0.8g/kg/d)|Placebo|3M|SOPS total score, positive, negative, disorganisation, general symptoms, depressive symptom severity, cognition|Glycine (Depressive symptoms)|Glycine (SOPS total, positive, negative, disorganisation, general symptoms, depressive symptoms)|

(Continued)

Table 7. Continued.

|Author, year/RCT/RoB|Centre|Referrals|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment#|Improve from baseline within group| |---|---|---|---|---|---|---|---|---|---|---|---|---| |(Woods S et al. 2017) Low RoB|Atlanta, Los Angeles, San Diego, San Francisco, Glen Oakes, Detroit, Worcester, Boston, New Haven, Chapel Hill, US, Calgary, Canada|Clinician referral|SIPS SOPS|51 24/27|22.3 16 to 40|DB|Ziprasidone (20-160mg/d)|Placebo|6M|Transition, positive symptoms|Ziprasidone (Positive symptoms)|NR| |(Zhu H et al. 2023) ChICTRI-OR-17013513 High RoB|Beijing Anding Hospital, China|NR|SIPS|80 37/40|19.76 16 to 30|DB|Shi-Zhen-An-Shen Herbal Formula Granule (SZAS-HFG) + placebo for antipirazole|Aripiprazole + placebo for SZAS-HFG|3M|Positive symptoms, cognitive functioning, general functioning|Shi-Zhen-An-Shen Herbal Formula Granule (SZAS-HFG) + placebo for aripiprazole (general functioning)|NR| |Combined (psychosocial plus pharmacological interventions (Bechdolf et al. 2011; 2017; 2023) osf = 631fa13b1e fB4d7686bf185 e6d89de17 Some concerns|6 sites in Germany (University of Cologne; Rheinlich-Wilhelm University; Heinrich-Heine University; Ludwig-Maximilian-Universität; Charte-Universitätsmedizin, Berlin; Campus Charte- Mitte, Berlin; University of Göttingen; University of Aachen)|Mental health professionals or self-referrals|SIPS SOPS SPIA|280 129/96/55|24.4 18 to 49|DB aripiprazole SB CBT|1,CBT 2. Aripiprazole + clinical management|Placebo + Clinical management|12M|Transition to psychosis, time to transition, psychopathological symptoms, psychosocial functioning, safety, side effects|None|All arms (Positive and disorganisation symptoms, basic symptoms (COGDS) psychosocial functioning, and depression)| |(McGorry et al. 2002) High RoB|PACE clinic, Melbourne, Australia|Mental health specialists, primary care, general population (schools)|CAARMS|59 31/28|20 14 to 30|SB|Specific Prevention Intervention (NIBl + risperidone [1-2mg] + CBT38)|NBI (pharmacological (AD, BOD2), psychological, psychosocial interventions including case-management with social support)|6M|Transition|SPI (transition)|NR| |(Yung et al. 2011; McGorry et al. 2013) ACTRN01260 5000247673 Low RoB|PACE Clinic, Melbourne, Australia|NR|CAARMS|115 43/44/28|18.1 14-30|DB|Risperidone (0.5-2mg/day) + CBT|CBT + placebo, supportive therapy + placebo|12M|Transition, psychiatric symptoms, psychosocial functioning, quality of life|None|Risperidone (0.5-2 mg/day) + CBT (negative symptoms, psychosocial functioning)|

(Continued)

|Author, year/RCT/RAoB|Centre|Referrals|CHR-P definition|N|Age|Blinding|Intervention|Control|Duration*|Primary and other outcomes|Superior treatment*|Improve from baseline within group| |---|---|---|---|---|---|---|---|---|---|---|---|---| |(MGorry et al., 2017; Nelson et al., 2018; Berger et al., 2019; Bob et al., 2019; Bolt et al., 2019; Bolt et al., 2019; Alqari et al., 2020; Amminger et al., 2020; Amminger, Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthamansachchi et al., 2020; Muthama|||||||||||||

24 M. SOLMI ET AL.

Bechdolf et al. 2023; McGorry et al. 2023; Wasserthal et al. 2024). The results are descriptively reported in Table 7 for both primary and secondary outcomes. Eight RCTs compared CBT with a control intervention, focusing on transition to psychosis as an outcome in six (Morrison et al. 2004; 2007; Addington et al. 2011; Morrison et al. 2012; van der Gaag et al. 2012; Ising et al. 2016; Stain et al. 2016; Pozza et al. 2019; 2021) of these. In only three (Morrison et al. 2004; 2007; van der Gaag et al. 2012; Ising et al. 2016; Pozza et al. 2019, 2021) of these six RCTs CBT outperformed the control group in reducing the transition rate. Of the 14 RCTs on other psychosocial interventions, only integrated psychological intervention outperformed the control group in two RCTs regarding transition to psychosis (Nordentoft et al. 2006; Bechdolf et al. 2012; Wessels et al. 2019), and one more recent RCT found no superiority for this intervention (Albert et al. 2017). Of the RCTs testing biological and combined interventions, $\omega$-3 PUFAs outperformed placebo on transition in one RCT (Amminger et al. 2010; Mossaheb, Papageorgiou, et al. 2018; Mossaheb, Schäfer, et al. 2018), but no difference emerged in three subsequent RCTs (Cadenhead K et al. 2017; McGorry et al. 2017; Nelson et al. 2018; Berger et al. 2019; Bolt et al. 2019; Cadenhead KS et al. 2019; Alqarni et al. 2020; Amminger et al. 2020; Mallawaarachchi et al. 2020; Qurashi et al. 2024). Of the five RCTs testing combined interventions, needs-based interventions plus low-dose risperidone and CBT outperformed needs-based intervention alone on transition in an earlier RCT (McGorry et al. 2002); however, four more recent RCTs found no superiority for any combined intervention over control conditions (Bechdolf et al. 2011; Yung et al. 2011; McGorry et al. 2013; Bechdolf et al. 2017; McGorry et al. 2017; Nelson et al. 2018; Berger et al. 2019; Bolt et al. 2019; Alqarni et al. 2020; Amminger et al. 2020; Mallawaarachchi et al. 2020; Bechdolf et al. 2023; McGorry et al. 2023; Wasserthal et al. 2024). No other psychosocial or biological or combined interventions was superior to other interventions on transition to psychosis in CHR-P subjects.

3.1.7. Meta-analyses and network meta-analyses of randomised controlled trials

The study selection flow of the umbrella review of (network) meta-analyses of RCTs in subjects meeting CHR-P status criteria, is reported in eFigure 7 (Table 8). Out of initial 350 hits, we ultimately included 11 (network) meta-analyses (Preti and Cella 2010; Stafford et al. 2013; van der Gaag et al. 2013; Hutton and Taylor

2014; Davies, Cipriani, et al. 2018; Davies, Radua, et al. 2018; Bosnjak Kuharic et al. 2019; Fusar-Poli, Radua, et al. 2021; Mei et al. 2021; Fusar-Poli, Radua, et al. 2022; Zheng et al. 2022). Studies excluded after full-text assessment, with reason for exclusion, are reported in eTable 9.

Details on outcomes, sample size, interventions, and results of included (network) meta-analyses are reported in Table 8.

Out of 11 publications, nine only conducted pairwise meta-analyses (Preti and Cella 2010; Stafford et al. 2013; van der Gaag et al. 2013; Hutton and Taylor 2014; Bosnjak Kuharic et al. 2019; Fusar-Poli, Radua, et al. 2021; Mei et al. 2021; Fusar-Poli, Radua, et al. 2022; Zheng et al. 2022), and two conducted a network meta-analysis (Davies, Cipriani, et al. 2018; Davies, Radua, et al. 2018). All (network) meta-analyses considered a mix of psychosocial and pharmacological interventions.

Among the two network meta-analyses, as primary outcomes, one considered psychotic symptoms (Davies, Radua, et al. 2018), and one considered transition to psychosis (Davies, Cipriani, et al. 2018). No significant result emerged for transition to psychosis. Ziprasidone as augmentation of needs-based intervention (NBI) outperformed family CBT with/without risperidone in one network meta-analysis (Davies, Radua, et al. 2018) regarding attenuated psychotic symptoms.

Regarding transition to psychosis, in pairwise meta-analyses, all interventions pooled together outperformed TAU in three meta-analyses (Preti and Cella 2010; van der Gaag et al. 2013; Mei et al. 2021). Antipsychotics outperformed TAU in one meta-analysis (van der Gaag et al. 2013). CBT outperformed supportive therapy in one meta-analysis (Stafford et al. 2013), and CBT plus supportive therapy outperformed supportive therapy alone in one meta-analysis (Bosnjak Kuharic et al. 2019). CBT-informed treatment outperformed other treatments in three meta-analyses (Hutton and Taylor 2014; Mei et al. 2021; Zheng et al. 2022). Importantly, significant results regarding CBT on transition to psychosis were not confirmed in any network meta-analysis (see above), nor in two additional meta-analyses (Fusar-Poli, Radua, et al. 2021; 2022). Regarding other outcomes, CBT also outperformed supportive therapy on psychotic symptoms in one meta-analysis (Hutton and Taylor 2014). No difference emerged from other outcomes (Hutton and Taylor 2014; Mei et al. 2021; Zheng et al. 2022). Differences in the included studies across eligible meta-analyses might explain (part of) the differences in results, are described in Table 8.

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

25

Table 8. Network and pairwise meta-analyses on randomised controlled trials aiming to prevent psychosis in people at clinical high risk of psychosis (expanded from (Fusar-Poli, Davies, et al. 2019)).

|Author, year|Source|Primary outcome, other outcomes|RCT-N|Interventions|Significant findings of primary analyses, random-effects model| |---|---|---|---|---|---| |(Preti and Cella 2010)|MA|Transition|5-18011-1,246|CBT, CBT+RIS, IPI, monitoring, NBI, OLA, Om3, Pbo, TAU|Transition: all interventions pooled together No other significant findings| |(Stafford et al. 2013)|MA|Transition||CBT, CBT+Pbo, CBT+RIS, CBT+SUP, IPI, NBI, NBI+AMI, OLA, Om3, Pbo, SUP, SUP+Pbo, TAU|Transition: i) CBT&gt;SUP at 6-12 months (RR = 0.54, 95%CI = 0.34-0.86), 12+ months (RR = 0.63, 95%CI = 0.40-0.99), ii) CBT+RIS&gt;SUP at 6 months (RR = 0.35, 95%CI = 0.13-0.95)| |(van der Gaag et al. 2013)|MA|Transition|10-NA|CBT, CBT+RIS, OLA, IPI, Pbo, TAU, antipsychotics|No other significant findingsTransition: i) all interventions pooled together, ii) Antipsychotics (RR = 0.55, 95%CI = NA, p = 0.029).| |(Hutton and Taylor 2014)|MA|Transition, APS, functioning, distress, QoL|6-800|CBT, CBT+monitoring, CBT+Pbo, CBT+RIS, IPI, monitoring, SUP, SUP+Pbo|Transition: CBT-formed &gt;others at 12 months (RR = 0.45, 95%CI = 0.28-0.73), at 18-24 months (RR = 0.41, 95%CI = 0.23-0.72).APS: CBT-informed &gt;others at 12 months (g=-0.25, 95%CI = -0.46to - 0.03)| |(Davies, Cipriani, et al. 2018)|NMA|Transition, acceptability|16-2,035|NBI, CBT-F + RIS+NBI, CBT-F + NBI, CBT-F + CBT-V + NBI, Dser+NBI, IPI, FFT+NBI, OLA+NBI, Om3+NBI, ZIP+NBI|No other significant findingsNo significant findings| |(Davies, Radua, et al. 2018)|NMA|APS, acceptability|14-1,707|NBI, ARI+NBI, CBT-F + NBI, CBT-F + RIS+NBI, Dser+NBI, FFT+NBI, IPI, OLA+NBI, Om3+NBI, ZIP+NBI|APS: i) ZIP+NBI&gt;NBI (SMD=-1.10, 95%CI=-2.04to - 0.15), &gt;CBT-F (SMD=-1.03, 95%CI=-2.05to - 0.01), &gt;CBT-F + RIS+NBI (SMD=-1.18, 95%CI -2.29to - 0.07) at 6 months.Acceptability: NBI+ARI&gt;NBI+OLA (OR = 3.73, 95%CI = 1.01-13.81)No other significant findings| |(Bosnjak Kuharic et al. 2019)|MA|Transition|20-2,151|NBI, Pbo, SUP, aminoacids, AMI+NBI, CBT, CBT+NBI + RIS, CBT+Pbo, CBT+RIS, CBT+SUP, CT, EC, FT, IPI, OLA + SUP, Om3, SUP, SUP+Pbo, TAU|Transition: CBT+SUP &gt;SUP transition 18 months (RR = 0.45, 95%CI = 0.23-0.89)No other significant findings| |(Mei et al. 2021)#|MA|Transition, APS, negative symptoms, depressive symptoms, anxiety symptoms, distress, QoL, functioning|26-2,351|CBD, CBCM+Pbo, CBT, CBT+monitoring, CBT+RIS, CBT+TAU, COMP, CRT, CRT+TAU, Dser, EC, FFT, FT, Gly, IPI, monitoring, NBI, NDRL+TAU, OLA+SUP, Om3, OBCM, Pbo, SUP, SUP+Pbo, TAU, ZIP|Transition: All interventions pooled together CBT&gt;others at EoT (RR = 0.55, 95%CI = 0.31-0.99), 12 months (RR = 0.52, 95%CI = 0.33-0.82), 18+ months (RR = 0.60, 95%CI = 0.42-0.84#)APS: All interventions pooled together No other significant findings| |(Fusar-Poli, Radua, et al. 2021)#|MA|Transition|7-NA|CBT, CBT+monitoring, CBT+Pbo, CBT+TAU, monitoring, NDRL+TAU, SUP, SUP+Pbo, TAU|No significant findings #| |(Zheng et al. 2022)*|MA|Transition, APS, depressive symptoms, distress, functioning, QoL|10-1,128|CBT, CBT+RIS, CBT+NBI, NBI|Transition: CBT&gt;NBI 6-12 (RR = 0.44, 95%CI = 0.30-0.64), 12-24 (RR = 0.46, 95%CI = 0.30-0.69), 24+ (RR = 0.58, 0.35-0.95) monthsNo other significant findings| |(Fusar-Poli, Radua, et al. 2022)##|MA|Transition|9-NA|CBT, CBT+RIS, CBT+NBI, NBI|No significant findings#|

Legend. Acceptability: all-cause discontinuation; AMl: amisulpride; APS: attenuated psychotic symptoms; ARl: aripiprazole; CBCM: CBCM, cognitive-behavioural case management; CBT: cognitive behavioural therapy; CBT-F: French & Morrison CBT protocol; CBT-V: van der Gaag CBT protocol; Cl: confidence interval; COMP: computer games; CT: cognitive training; Dser; D-serine; EC: enhanced care; FFT: family-focused therapy; FT: family treatment; Gly: glycine; IPL: integrated psychological interventions; MA: meta-analysis; NBl: needs-based interventions (including placebo); NDRL: non directive reflective listening; NMA: network meta-analysis; OLA: olanzapine; Om3: omega-3 fatty acids; Pbo: placebo; QoL: quality of life; RCT: randomised controlled trial; RIS: risperidone; RR: risk ratio; SUP: supportive; TAU: treatment as usual; Total RCT/n: overall number of RCTs and subjects included in the meta-analysis or network meta-analysis across any outcome; ZIP: ziprasidone. Bold outcome was primary outcome of review; +: GRADE=very low; +: excluding PREVENT, and including RCT with high risk of bias; +: excluding RCT with high risk of bias, and including PREVENT trial; +: excluding PREVENT, including three RCTs; +: excluding RCT with high risk of bias, one Chinese RCT with high risk of bias, and including PREVENT trial; +: excluding PREVENT, including three RCTs with high risk of bias. Edited from (Preti and Cella 2010; Stafford et al. 2013; van der Gaag et al. 2013; Hutton and Taylor 2014; Schmidt SJ et al. 2015; Davies, Cipriani, et al. 2018; Davies, Radua, et al. 2018; Devoe et al. 2018; Bosnjak Kuharic et al. 2019; Devoe et al. 2019b, 2019a, 2020; Fusar-Poli, Radua, et al. 2021; Mei et al. 2021; Fusar-Poli, Radua, et al. 2022; Zheng et al. 2022).

26

M. SOLMI ET AL.

3.1.8. Studies comparing psychosis prevention services with other services, and cost-effectiveness of psychosis prevention services

We updated a systematic review (Sizer et al. 2022) on studies comparing psychosis prevention services with other services (Table 9). The study selection flow of the update of that systematic review is reported in eFigure 8. Studies excluded after full-text assessment,

with reason for exclusion, are reported in eTable 10. Three cohort studies were reported in the original systematic review (Sizer et al. 2022), comparing 78 subjects with FEP after meeting CHR-P criteria in prevention services (CHR-P-T) versus 253 subjects not being assessed in psychosis prevention services and directly presenting with FEP (FEP-D) (Table 9). One study set in Taiwan showed no difference on cognitive

Table 9. Studies comparing psychosis prevention services with other services, and cost-effectiveness of psychosis prevention services.

|Author, yearCountry|NMonths|DesignModelIntervention/service|Outcomes|Findings| |---|---|---|---|---| |Psychosis prevention (Yoviene Sykes et al. 2020)|9712M|CohortMean or frequency comparisonsFEP seen in UHR service vs FEPnot seen|Any hospitalisation, number of hospitalisations, length of stay, any emergency room visit, repeat diagnosis of psychosis or continued antipsychotic prescription, DUP|UHR service superior in DUPUHR had more emergency visits.No difference in any hospitalisation, number of hospitalisations, length of hospital stay, repeat diagnosis of psychosis.| |(Valmaggia et al. 2015)|19012M|CohortMean or frequency comparisonsFEP seen in UHR service vs FEPnot seen|Any hospitalisation, total involuntary admissions, length of hospital stay, DUP|UHR service superior in any hospitalisation, involuntary admissions, DUP.No difference in length of hospital stayNo difference in any cognitive function| |(Liu et al. 2015)|6712M|CohortMean or frequency comparisonsFEP seen in UHR service vs FEPnot seen|WCST, WAIS-III, TMT, verbal fluency test, WeschlerMemory test|| |(Fusar-Poli P. et al. 2020)|156124M|CohortMean or frequency comparisonsFEP seen in UHR service vs FEPnot seen|Treatment with antipsychotic, informal hospitalisation, compulsory hospitalisation, treatment with clozapine, and length in hospital stay|UHR service inferior in antipsychotic medication (at 3, 6, and 24 months, not 12months), informal and involuntary hospital admission (at all time points), and length in hospital stay (at all time points).No difference in clozapine| |Cost-effectiveness(Ising et al. 2015)|19618M|RCTIncremental cost-effectiveness ratio with nonparametric bootstrappingTAU+CBT vs TAU|Transition, QALY.Intervention, medical, travel, loss of productivity costs.|Cost-effectiveness: CBT avoided 89.5% v. 76.2% transition to psychosis, lower cost by US$8844 (£551) per prevented psychosis, 63.7% probability of being more cost effective.Cost-utility: CBT had similar QALY 0.60 v. 0.57, 52.3% probability of being the superior treatment.| |NL|19648M|RCTIncremental cost-effectiveness ratio with nonparametric bootstrappingTAU+CBT vs TAU|Transition, QALY.Intervention, medical, travel, loss of productivity costs.|Cost-effectiveness: CBT avoided transition to psychosis (with a risk difference of 0.122; b=1.324, SEb = 0.017, z=7.99, p&lt;0.001), lower cost by US$ 577 per participant, 83% probability of being more cost effective.Cost-utility: CBT had similar QALY 2.63 vs 2.46, 75% probability of being superior treatment, and 92% probability of being cost-effective at €20 000 threshold per QALY.| |(Ologundudu et al. 2023)|100015year|Cohort (simulated)Decision modelling, Markov modelStandard treat all CHR-P strategy vs Risk stratification treatment strategy (i.e. CHR-P with ≥20 % risk of transition only)|Transition, QALY.|Treating only CHR-P individuals with a risk of ≥20% resulted in lower costs ($1,398) and fewer quality-adjusted life-years (0.055 QALYs) per person compared to a strategy of treating every CHR-P individual. The incremental cost-effectiveness ratio (ICER) for the &#x27;treat all&#x27; approach was $25,448 per QALY, indicating that treating all individuals may be a cost-effective strategy.|

(Continued)

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

27

Table 9. Continued.

|Author, yearCountry|NMonths|DesignModelIntervention/service|Outcomes|Findings| |---|---|---|---|---| |(Phillips et al. 2009)|59|RCT|Transition to psychosis.|Lower transition to psychosis, but no| |AU|36M|Mean costs comparison|Outpatient, pharmacological,|difference in costs| |(Perez et al. 2015)|104 centres|SPI vs NBI|patient costs.|| |UK|24M|Cluster RCT|Number of referrals, true|Total cost per true-positive referral in the 2 year follow-up was £26 785 in high-intensity practices, £27 840 in low-intensity practices, and £30 007 in PAU practices. The high-intensive intervention, which improved liaison between primary and secondary care for people presenting early signs of psychosis, was found to be both clinically and cost effective.| |(Valmaggia et al. 2009)|114|Cohort|Services use and|Over the initial 12 months from presentation, the costs of the OASIS intervention were £1872 higher than TAU. However, after 24 months they were £961 less than TAU.| |UK|24M|Decision modelling|pharmacological treatment|presentation, the costs of the OASIS intervention were £1872 higher than TAU. However, after 24 months they were £961 less than TAU.| |(Wijnen et al. 2020)|2,514|Cohort (simulated)|Service use, treatment and utilities|CBTuhr resulted in a per-patient increase of 0.06 QALYs and a per patient cost reduction of €654 (dominant incremental cost-effectiveness ratio) with a reduction in 5-year healthcare costs of €1,002,166.| |NL|120M|PsyMod, Markov model|National health service costs, productivity costs|Reach all new CHR-P cases each year would need extra costs of £1 904 396 in first year (from increased service costs), but free up resources of £19.7m over 10 years with an estimated 10-year cost of implementation gap for each 1 year cohort of £18.9m| |(Campion et al. 2019)|857/year|Cohort (model)|National health service costs, productivity costs|Mental health improved or stayed the same, employment rates improved, total cost difference of £621 425| |UK|120M|Knapp model|Psychosis prevention service|Cost-effective above willingness-to-pay £20 000 and £30 000 threshold per QALY| |||(OASIS) vs standard care||| |(Brimblecombe et al. 2017)|20|Cohort|Service use, crime, employment|Cost-effective above willingness-to-pay £20 000 and £30 000 threshold per QALY| |UK|36M|Pre vs post|costs|| |||Psychosis prevention service, within-subject pre-post comparison||| |(Jin et al. 2021)|200,000|Cohort (simulated)|Lifetime costs, QALY|Cost-effective above willingness-to-pay £20 000 and £30 000 threshold per QALY| |UK|Lifetime|Cost-utility analysis, discrete-event simulation|Cost UHR 100% vs 41.01%|Cost-effective above willingness-to-pay £20 000 threshold per QALY| |(Jin et al. 2020)|69,800|Cohort (simulated)|Clinical benefits, clinical harms, costs, cost savings|| |UK|Lifetime|Cost-utility analysis, discrete-event simulation||| |||TAU + CBT vs TAU|||

Legend. CBT: cognitive behavioural therapy; DUP: duration of untreated psychosis; FEP: first-episode psychosis; NBl: supportive and ass needed psychotherapy; OASIS: Outreach and Support in South London; SPI: specific preventive intervention, including pharmacological treatment with antipsychotics and psychological treatment; TRS: treatment-resistant schizophrenia; UHR: ultra-high risk of psychosis; WCST: Wisconsin Card Sorting Test; WAIS-III: Wechsler Adult Intelligence Scale-3rd Edition; TMT: Trail Making Tests. Edited from (Phillips et al. 2009; Valmaggia et al. 2009; Ising et al. 2015; Liu et al. 2015; Perez et al. 2015; Valmaggia et al. 2015; Brimblecombe et al. 2017; Ising et al. 2017; Campion et al. 2019; Fusar-Poli P. et al. 2020; Jin et al. 2020; Wijnen et al. 2020; Yoviene Sykes et al. 2020; Jin et al. 2021; Sizer et al. 2022; Ologundudu et al. 2023).

functioning, suggesting that the access to CHR-P and related help-seeking behaviour before FEP might not be driven by cognitive symptoms (Liu et al. 2015). Regarding outcomes after FEP, one study set in UK showed that CHR-P-T subjects had fewer hospitalisations, fewer involuntary admissions, and shorter duration of untreated illness (Valmaggia et al. 2015). This latter finding was confirmed in one US study (Yoviene Sykes et al. 2020). However, no differences emerged on the length of hospital stay in both the UK and US, and in the US no differences emerged for hospitalisations. We identified one additional study (Fusar-Poli P.

et al. 2020) set in UK showing that CHR-P-T subjects were more likely to receive antipsychotic medication (at 3, 6, and 24 months, not at 12 months), be admitted to hospital both informally and on a compulsory basis, and spend more time in hospital compared to FEP-D subjects, highlighting severe clinical outcomes for CHR-P-T. No differences were observed in the likelihood of receiving clozapine.

Regarding cost-effectiveness, the study selection flow of the systematic review of cohort and cost-effectiveness studies regarding psychosis prevention services is reported in eFigure 9. Studies excluded after full-text assessment,

28

M. SOLMI ET AL.

with reason for exclusion, are reported in eTable 11. From initial 2956 hits we ultimately included 11 studies (Phillips et al. 2009; Valmaggia et al. 2009; Ising et al. 2015; Perez et al. 2015; Brimblecombe et al. 2017; Ising et al. 2017; Campion et al. 2019; Jin et al. 2020; Wijnen et al. 2020; Jin et al. 2021; Ologundudu et al. 2023). Details and results of included studies are reported in Table 9. Six studies were set in UK, three in the Netherlands, one in Australia, and one in Canada. Five studies used data that originated from a single RCT comparing CBT with TAU, and simulated cost-effectiveness models of CBT confirming its cost-effectiveness (Ising et al. 2015; 2017; Jin et al. 2020; Wijnen et al. 2020; Jin et al. 2021). One study measured actual costs associated with specific preventive intervention (SPI) versus need-based intervention (NBI), instead showing no difference in costs (Phillips et al. 2009). Another study found that high-intensive interventions strengthening liaison between primary and secondary care for people presenting with early signs of psychosis was cost-effective (Phillips et al. 2009). Two other studies used real-world data from the OASIS psychosis prevention service, London, UK, showing its cost-effectiveness versus standard care (Valmaggia et al. 2009; Campion et al. 2019). One further study measured the costs before versus after receiving an intervention in a psychosis prevention service, again showing cost effectiveness (Brimblecombe et al. 2017). Finally, a simulation study demonstrated that providing treatment to all CHR-P individuals, regardless of their transition risk as assessed by the North American Prodrome Longitudinal Study risk calculator, led to higher quality-adjusted life-years (QALYs) compared to a risk-stratified approach and was found to be cost-effective (Ologundudu et al. 2023).

3.2. Grading, guidelines, and recommendations

3.2.1. Grading of evidence

The criteria used to grade the level of evidence, and formulate guidelines are available in Tables 1 and 10.

Grading of evidence and guidelines for prevention of psychosis at intervention- and service-levels are available in Table 11. Table 11 is limited to those interventions with $\ge 1$ study showing the intervention or the service to be superior to the control group on transition to psychosis.

3.2.2. Recommendations

Considering the evidence and its grading an initial list of 12 recommendations was circulated to authors. Scores were ≥4/5 in ≥ 80% of respondents for all 12 recommendations. The recommendations are listed in Table 12 and reported below together with the

consensus rates from authors, grouped by service organisation, assessment and risk communication, and intervention level.

Three recommendations were made at service organisation level:

1. Psychosis indicated prevention services might be implemented in close collaboration with early intervention services for psychosis to minimise duration of untreated illness. 90.7% agreed 4 or 5/5.

2. To prevent psychosis, at the mental health service level, intake age criteria should be between 14 to 35. 87.5% agreed 4 or 5/5.

3. To prevent psychosis, mental health services should allow access to persons with cannabis use disorder. 96.9% agreed 4 or 5/5.

Six recommendations were made at the assessment and risk communication level:

1. To identify CHR-P, mental health services staff should be trained to administering and rating CAARMS, SIPS, or future validated interviews such as PSYCHS in clinical settings. 84.4% agreed 4 or 5/5.

2. To identify CHR-P, mental health services staff should use CAARMS, SIPS, or PSYCHS in the future in clinical settings. 84.4% agreed 4 or 5/5.

3. To estimate risk of psychosis, mental health services staff should be trained in using and interpreting psychosis risk calculators in clinical settings. 81.3% agreed 4 or 5/5.

4. To communicate the risk of psychosis, mental health services staff should be trained in communicating the estimate of risk of psychosis in clinical settings. 84.4% agreed 4 or 5/5.

5. To estimate and communicate the risk of psychosis, mental health services staff should use only validated risk calculators, keeping a human-to-human interaction with client. 81.3% agreed 4 or 5/5.

6. To provide comprehensive care, mental health services staff should assess the presence of comorbid mental disorders also in those meeting CHR-P criteria. 100% agreed 4 or 5/5.

Three recommendations were made at intervention level:

1. To prevent psychosis, mental health services staff should offer treatment to reduce use or achieve abstinence of cannabis use (if present). 90.7% agreed 4 or 5/5.

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

29

Table 10. World federation of societies of biological Psychiatry grading system for levels of evidence.

|Evidence effective/ Grade|Explanation|Evidence not effective/ Grade|Explanation| |---|---|---|---| |Strong/ A|At least two independent RCTs with a low risk of bias show efficacy (superiority to placebo, or in the case of psychotherapy studies, superiority to a &#x27;active psychological placebo&#x27; in a study with adequate blinding), OR superiority to/equivalent efficacy compared with an established comparator treatment in a 3-arm study with placebo control or in a well-powered non-inferiority trial (only applicable if such a standard treatment exists) with a low risk of bias, AND No negative RCTs with a low risk of bias exist. If there are contradicting results from RCTs, the majority of RCTs AND/OR a meta-analysis with low risk of bias shows efficacy. If there are more than one &#x27;A&#x27; treatment options, the decision should be based on head-to-head comparisons or meta-analyses showing superiority of one of the treatments|Strong/ -A|At least two independent, adequately powered RCTs with low risk of bias as detailed on the left show NO efficacy AND no positive RCTs with a low risk of bias exist If there are contradicting results from RCTs, however, the majority of RCTs AND/OR a meta-analysis with very low risk of bias shows NO efficacy| |Limited/ B|RCT with a moderate risk of bias showing superiority to placebo (or in the case of psychotherapy studies, superiority to a &#x27;active psychological placebo&#x27;) OR a randomised controlled comparison with a standard treatment without placebo control with a sample size sufficient for a non-inferiority trial with a moderate risk of bias, AND No negative studies exist OR Meta-Analyses with a moderate risk of bias that shows efficacy|Limited/ -B|1 RCT with a moderate risk of bias showing NO superiority to placebo (or in the case of psychotherapy studies, NO superiority to a &#x27;active psychological placebo&#x27;) OR LESS efficacy than a standard treatment in an RCT without placebo control with a moderate risk of bias AND No positive studies exist OR Meta-Analyses with a moderate risk of bias that show NO efficacy| |Low/ C1|One or more prospective open studies (with a minimum of 10 evaluable patients per group) using a control group, but no randomisation, or using no control group, show efficacy. OR one or more well-conducted case control or cohort studies (with a minimum of 10 evaluable patients) with a moderate probability that the relationship is causal show efficacy OR RCTs AND/OR meta-analyses with a high risk of bias show efficacy|Low/ -C1|One or more prospective open studies (with a minimum of 10 evaluable patients) using a control group, but no randomisation, or using no control group, show NO efficacy. OR One or more well-conducted case control or cohort studies (with a minimum of 10 evaluable patients) with a moderate probability that the relationship is causal shows NO efficacy OR RCTS AND/OR meta-Analyses with a high risk of bias that show NO efficacy| |Low/ C2|Non-analytic studies, e.g. case reports or case series with less than 10 evaluable patients show efficacy in the majority of cases|Low/ -C2|Non-analytic studies, e.g. case reports or case series with less than 10 evaluable patients show NO efficacy in the majority of cases| |Low/ C3|Expert opinions not based on any published data reporting efficacy|Low/ -C3|Expert opinions not based on any published data reporting NO efficacy| |No evidence|No sufficient evidence to advise for or against the use of the intervention|No evidence|No sufficient evidence to advise for or against the use of the intervention|

Edited from (Hasan et al. 2019).

2. To provide comprehensive care, mental health services staff should offer evidence-based treatment for comorbid mental disorders also in those meeting CHR-P criteria. 100% agreed 4 or 5/5.

3. To prevent psychosis, mental health services staff should offer treatment for CHR-P status based on patient preference, clinician/centre experience, and following the ‘first do no harm’ principle. 93.8% agreed 4 or 5/5.

4. Discussion

These guidelines update previous guidelines, in light of updated evidence (NICE 2014; Schmidt SJ et al. 2015; Addington et al. 2017), integrating a lived experience and clinical experience-informed a-priori-defined lens that is representative of the clinical and ethical matters related with the indicated prevention of psychosis. All specific interventions including CBT, other psychosocial interventions, biological interventions and combined

30 M. SOLMI ET AL.

Table 11. Guidelines for the prevention of psychosis from the World Federation of Societies of Biological Psychiatry and EPI Canada, on interventions with at least one positive randomised controlled trial on transition to psychosis in clinical high risk of psychosis subjects, or at least one study showing cost-effectiveness.

|Intervention|LoE primary outcome|LoE key and other secondary outcomes|Acceptability|Recommendation| |---|---|---|---|---| |Cognitive therapy or CBT for psychosis (k=7)|No recommendations RCT:3 superior, high RoB3 negative, high RoBMA RCT:6MA superior, high RoB2MA negative, some RoBNMA RCT:1 negative, high RoB|No recommendations RCT:3 superior (depressive, anxiety, insomnia symptoms), high/low RoB7 negative, high/low RoB1 inferior (distress), high RoBMA:1 superior, high RoB2 negative, high RoBNMA:1 inferior, high RoB|GOOD5 cost effective|No recommendation| |Integrated treatment (k=2)|No recommendations RCT:2 superior, high RoB1 negative, high RoBMA RCT:3 negative, high RoB3 inferior, high RoBNMA RCT:1 negative, high RoB|No recommendations RCT:1 superior (negative symptoms), high RoB2 negative, high RoBMAA:2 negative, high RoBNMA:1 negative, high RoB|NR|No recommendation| |Risperidone (including SPI) (k=2)|No recommendations RCT:1 superior, high RoB1 negative, low RoBMA RCT:5 superior, high RoB3 negative, high RoBNMA RCT:1 negative, high RoB|No recommendations RCT:1 negative, low RoB1 NAM:2 negative, high RoBNMA:1 inferior, high RoB|POORNot cost effective (SPI)|No recommendation| |ω-3 PUFAs (k=3)|No recommendations RCT:1 superior, low RoB3 negative, low RoBMA RCT:3 negative, high Rob2 inferior, high RoBNMA RCT:1 negative, high RoB|No recommendations RCT:1 positive (positive, negative, general symptoms, functioning), low RoB2 negative, low RoB1 inferior (functioning), low RoBMA:3 negative, high RoBNMA:1 negative, high RoB|NR|No recommendation| |Psychosis prevention service|B MA RCT:3 positive pooling different interventions, and no superior intervention in NMA|C1Cohort:1 cohort: mental health improved or stayed the same, fewer hospitalisations, involuntary admission.2 cohort: DUP shortened2 negative: length of hospital stay1 negative: number of hospitalisations1 negative: clozapine1 inferior: informal and involuntary hospital admission, and length in hospital stay|GOOD3 cost-effective|Limited recommendation|

Legend. CBT: cognitive-behavioural therapy; LoE: level of evidence; MA: meta-analysis; NA: did not assess; NMA: network meta-analysis; NR: no studies reported; RCT: randomised controlled trial; RoB: risk of bias; SPI: specific prevention intervention.

interventions, had conflicting findings at the individual RCT and meta-analytic level, precluding any possibility of recommending specific interventions for the prevention of psychosis. Some psychological interventions, such as CBT, have been tested more extensively than others and are therefore more likely to be widely available. Evidence on CBT has been mixed, with some studies supporting its superiority to treatment as usual and

others not, including a meta-analysis published after our search (Minichino et al. 2025). Notably, specific interventions pooled together outperformed treatment as usual in several separate meta-analyses (Phillips et al. 2009; Preti and Cella 2010; van der Gaag et al. 2013). Moreover, observational evidence showed that psychosis prevention services improved several outcomes compared with other services (Valmaggia et al. 2015; Yoviene

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

31

Table 12. Clinical recommendations for the prevention of psychosis.

|Recommendations|Evidence| |---|---| |Service organisation|| |Psychosis indicated prevention services might be implemented in close collaboration with early intervention services for psychosis to minimise duration of untreated illness|Observational| |To prevent psychosis, at the mental health service level, intake age criteria should be between 14 to 35.|Observational| |To prevent psychosis, mental health services should allow access to persons with cannabis use disorder.|Observational| |Assessment and risk communication|| |To identify CHR-P, mental health services staff should be trained to administering and rating CAARMS, SIPS, or future validated interviews such as PSYCHS* in clinical settings.|Observational| |To identify CHR-P, mental health services staff should use CAARMS, SIPS, PSYCHS* in clinical settings.|Observational| |To estimate risk of psychosis, mental health services staff should be trained in using and interpreting psychosis risk calculators in clinical settings.|Consensus only| |To estimate and communicate the risk of psychosis, mental health services staff should be trained in communicating the estimate and risk of psychosis in clinical settings.|Consensus only| |To estimate and communicate the risk of psychosis, mental health services staff should use only validated risk calculators, keeping a human-to-human interaction with client.|Consensus only| |To provide comprehensive care, mental health services staff should assess the presence of comorbid mental disorders also in those meeting CHR-P criteria.|Observational| |Intervention|| |To prevent psychosis, mental health services staff should offer treatment aiming to abstinence from cannabis use (if present).|Observational| |To provide comprehensive care, mental health services staff should offer evidence-based treatment for comorbid mental disorders also in those meeting CHR-P criteria.|Observational| |To prevent psychosis and provide comprehensive care, mental health services staff should offer treatment for CHR-P status based on patient preference, clinician/centre experience, and following the &quot;first do no harm&quot; principle.|Randomised controlled trials|

Legend. *: based on future evidence of psychometric properties; CAARMS: Comprehensive Assessment of At-Risk Mental State; CHR-P: clinical high-risk for psychosis; PSYCHS: Positive SYmptoms for CAARMS Harmonised with SIPS; SIPS, Structured Interview for Prodromal Symptoms.

Sykes et al. 2020), and that prevention services, and CBT in particular, were cost-effective (Valmaggia et al. 2009; Ising et al. 2015; Perez et al. 2015; Brimblecombe et al. 2017; Ising et al. 2017; Campion et al. 2019; Jin et al. 2020; Wijnen et al. 2020; Jin et al. 2021). However, given the lack of consistent findings across individual RCTs, and different meta-analyses, these guidelines can make 'limited' recommendations to implement psychosis prevention services, and cannot recommend one specific intervention in individuals at CHR-P among many that have been previously tested.

While RCTs for schizophrenia select a potentially biased sample that is not necessarily representative of the general population (Taipale et al. 2022), there is no evidence of selection/sampling bias in RCTs of individuals at CHR-P for prevention of psychosis, as rates of transition to psychosis do not differ between cohort studies and RCTs (Salazar de Pablo, Davies, et al. 2021). Hence, findings from RCTs included in this review probably can be extended to real-world services. Psychosis prevention services are mainly designed to provide primary indicated prevention in those meeting CHR-P criteria. However, beyond indicated prevention, psychosis prevention services also implement several primary public health prevention strategies (e.g. universal and selective interventions) globally (Estradé et al. 2022). A systematic review including 66 publications on 13 standalone, 40 integrated, three network, and six regional or international surveys of CHR-P services across Europe, Asia, Oceania, Africa, North and South America, providing

care to >28 million people, showed that CHR-P services already target social and cultural (16 initiatives), economic (seven initiatives), demographic (six initiatives), environmental events (four initiatives) and neighbourhood (three initiatives) Sustainable Development Goal determinants of mental disorders developed by United Nations Member States (Estradé et al. 2022), indicating that CHR-P services deliver a wide range of public health initiatives aimed at prevention and the promotion of mental well-being, extending beyond the indicated prevention of psychosis. We believe that given the heterogeneous clinical practice across psychosis prevention services, pooling all interventions reflects a diverse portfolio of interventions that are different in quality and superior in intensity from treatment as usual delivered in 'regular' mental health services, which primarily focus on more advanced stages of persistent and severe mental disorders. However, there is the need for more research on individualised and adapted implementation of psychosis prevention services according to given cultural, environmental and resource parameters.

The specific assessment and intervention model to be administered within each psychosis prevention service should be informed by the clinical presentation of each patient, and should be informed by evidence on CHR-P assessment tools, risk factors for psychosis, psychosis risk modelling, concomitant mental disorders, and, importantly, ethical considerations. Some clinical recommendations were nevertheless made based on the evidence review and consensus.

32 M. SOLMI ET AL.

Mental health professionals working in psychosis prevention services should be trained in using the CAARMS or SIPS/SOPS to assess/determine CHR-P status. Recently a novel tool for CHR-P assessment is being tested - i.e. the PSYCHS (Woods SW et al. 2024) – to overcome limitations of CAARMS or SIPS/SOPS, and to provide one unique tool, which should be ideally be adopted once evidence is available regarding its psychometric properties and clinical usability (Brady and Larrauri 2023). It is important that the service organisation accounts for the need of administering such tools, which can be time-consuming but that are needed to define CHR-P. Moreover, abbreviated versions of CHR-P tools are now available such as the Mini-CAARMS (Stefana et al. 2025), reducing administration time.

Communicating risk estimates in medicine is complex. Risk of transition to psychosis can be calculated, yet this should be done exclusively using multivariable, validated, transparent and reproducible clinical prediction models. Moreover, risk models should always be applied in the context of a human-to-human interaction and by mental health professionals that are trained in clinical prediction model interpretation and communication. Overall, communication with patients and families about psychosis risk should include encouraging patients to state their preferences regarding the communication of medical information, respecting requests not to receive certain information, and tailoring disclosure to the individual's needs and expectations (Woods SW et al. 2021), consistently with ethical guidance from the American Medical Association (AMA n.d.). Any communication to patients regarding their risk should be conducted by staff trained in communicating risk, which they should be accountable for.

Cannabis use is a risk factor for schizophrenia, and other mental disorders (Myran, Harrison, et al. 2023; Myran, Pugliese, et al. 2023; Myran et al. 2024; Myran, Pugliese, Harrison, et al. 2025; Myran, Pugliese, McDonald, et al. 2025). It has been estimated that the population attributable fraction of cannabis with respect to schizophrenia spectrum disorders is 9.73% (95% CI = 4.50–17.30%) (Dragotiot et al. 2022). The risk is particularly elevated in young males, and progressively increases from persons with cannabis use disorder to cannabis-induced psychosis (Myran, Harrison, et al. 2023). Considering the age at onset of psychosis, and other mental disorders (Solmi, Radua, et al. 2022), young adults should be informed regarding the risk of cannabis. Unfortunately, the literacy regarding the risk of

psychosis associated with use of cannabis is very low (Goodman and Hammond 2022).

Several treatments have been tested for the treatment of cannabis use disorder. Evidence suggests that combining CBT, motivational enhancement therapy (which seeks to enhance motivation to change), and contingency management (which uses operant conditioning to reinforce a target behaviour) is effective in reducing frequency and quantity of use. However, improvements in sustained abstinence remain limited. In this context, testing novel pharmacological agents represents a promising avenue to enhance the effects of existing psychosocial treatments (Sherman and McRae-Clark 2016).

To prevent psychosis and avert adverse developmental effects of presenting symptoms, comorbidities and functional disabilities, psychosis prevention services should not leave out patients younger than 18. The peak age at onset of schizophrenia is 20.5 years, and a reasonable age intake criterion is age 14 to 35 also considering feasibility and resource allocation, as well as specificities of younger populations who frequently need care for different conditions (Solmi, Radua, et al. 2022). Indeed, the high rates of comorbid mental disorders in CHR-P call for a transdiagnostic mental health assessment. Individuals meeting CHR-P criteria should be assessed for and receive evidence-based treatment for concomitant mental disorders targeting also an improvement of functioning. The choice of specific interventions offered to individuals at CHR-P should be based on the person's autonomy and informed shared decision-making and consent process, on a 'first do no harm' non-maleficence principle, and on clinical expertise of a multi-professional mental health team. Regarding comorbid disorders, we could not recommend specific interventions to improve outcomes in CHR-P youth with specific comorbid mental health disorders due to the lack of meta-analytic evidence. Moreover, non-psychotic symptoms (e.g. depressive and anxiety symptoms) have not shown improvement in experimental interventions compared to control conditions in 'generic' samples (i.e. samples not stratified by comorbid mental disorders) of CHR-P individuals. Hence the treatment of comorbid mental disorder should follow evidence-based guidelines for each disorder.

Given that young people with mental health conditions tend to face substantial delays in accessing treat-ment (McGorry et al. 2024), it is essential to foster continuity of care and strengthen referral pathways, for example by promoting connections with emerging

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

33

youth-oriented primary care transdiagnostic services (McGorry et al. 2022).

However, risk estimates need to be communicated in even more cautious and balanced ways in younger subjects. The question whether psychosis prevention services or transdiagnostic mental health services should be implemented remains currently unanswered (Solmi, Soardo, et al. 2023).

Future research should address several aspects that could complement the evidence that informed these guidelines. First, clinicians, ethics experts, subjects with lived experience and their families should work together in co-designing a standardised approach to communicate risk of psychosis. Second, more research needs to be conducted regarding the impact of cannabis strength (i.e. THC content), frequency and quantity of use, on the risk of psychosis and on the outcome of psychosis (after onset) in subjects with or without CHR-P. Third, additional treatments should be conducted to prevent psychosis in CHR-P, testing novel biological pharmacological, non-pharmacological (e.g. brain stimulation), lifestyle (e.g. exercise, diet, smoking cessation), digital or psychosocial interventions. Future multi-arm, adaptive RCTs or implementation science approaches may represent a future direction in the field. Fourth, RCTs are needed to test candidate interventions to prevent psychosis in subjects with cannabis use disorder. Fifth, data from previous RCTs to prevent psychosis in CHR-P should be shared in collaborative efforts to analyse individual patient data to identify subgroups of patients that might benefit from a specific type of treatment. Sixth, while implementing cluster RCTs comparing psychosis prevention service versus transdiagnostic mental health prevention services would pose feasibility barriers, (simulation) studies leveraging data from existing services, and/or Delphi studies collecting the input of persons with lived experience in co-designing prevention services are needed. Seventh, our recommendations were not informed by neuroimaging and biomarker findings, as they are not yet informative for guiding real-world clinical care at scale (Rambaud et al. 2022) due to their small effect sizes, limited replication, lack of evidence testing their clinical utility, and poor scalability. Nevertheless, machine learning approaches and biomarker research may hold promise in advancing knowledge on heterogeneity among youth at CHR-P (Koutsouleris et al. 2021). Future studies employing multimodal and promptly available, scalable parameters may leverage neuroimaging data alongside behavioural and symptom data

to refine prognostic precision in this population, in line with ongoing large-scale international initiatives such as the Accelerating Medicines Partnership (AMP®) Schizophrenia (SCZ) program (Brady and Larrauri 2023; Wannan et al. 2024). Eighth, additional studies testing the prognostic accuracy and broader clinical utility of interviews measuring basic symptoms, potentially at even earlier stages than attenuated psychotic symptoms, accounting for language- and context-specific considerations, should be conducted (Schultze-Lutter et al. 2012). Ninth, research focusing on both psychosis prevention services and transdiagnostic prevention services is still limited. To date, tools to assess bipolar at-risk status are being developed (Salazar de Pablo et al. 2025; Stefanelli et al. 2025), as well as even broader approaches such as the clinical high at risk mental state (CHARMS) (McGorry et al. 2018; Destrée et al. 2024), and transdiagnostic prediction models are emerging (Arribas et al. 2026), yet no comparative study has analysed the clinical utility of a transdiagnostic versus a psychosis-focused prevention approach. Tenth, future research could focus on therapeutic monitoring of pharmacological treatment in CHR-P subjects who receive a pharmacological (non-antipsychotic) treatment for comorbid non-psychotic mental disorders. Such research could help develop and examine tools to optimise treatment personalisation. This is in line with recent research advances on therapeutic monitoring of antipsychotic medications in individuals with full-blown psychotic disorders (Hart et al. 2024; Schoretsanitis et al. 2024).

These guidelines have several limitations related with the restrictions of the currently available data, and of previous evidence synthesis efforts. For instance, the recommendation to target cannabis use was not based on evidence demonstrating that a reduction in cannabis use results in lower psychosis transition rates. It was based on the convincing evidence of cannabis as a modifiable risk factor for psychosis. Moreover, prevention in specific populations, such as first-degree relatives of persons with schizophrenia not meeting criteria for CHR-P, individuals with 22q11 syndrome, or those who have not been assessed with CHR-P tools or that do not have cannabis use disorder go beyond the aims and scope of these guidelines.

Nevertheless, we the current guidelines provide context, content, tools and specific guidance that is useful for researchers, clinicians, health care administrators and other interest who aim at reducing the risk for individuals to develop psychosis.

34

M. SOLMI ET AL.

Acknowledgements

MS is supported by Clinical Research Chair in Evidence-Based Mental Health, from the Faculty of Medicine, University of Ottawa, Canada. MB is supported by a NHMRC Senior Principal Research Fellowship and Leadership 3 Investigator grant (1156072 and 2017131). TB is supported by funding from the Ministry of University and Research under the call Progetti di Rilevante Interesse Nazionale (PRIN) 2022 (project number 20224SX547). JS is supported by a FRQS Senior Clinician-Scientist Award. Alison Yung is supported by an NHMRC leadership fellowship #2026339.

Author contributions

Conceptualisation: MS; Methodology: MS, CUC, AFC, TB, GLB; Writing - Original draft: MS, GLB, TB; Visualisation: MS, GLB, RG, TB; Investigation: MS, GLB, TB, RG; Supervision: MS, CUC, AFC; Project administration: MS; Writing - review & Editing: all authors

Disclosure statement

Consultant: AbbVie, Alkermes, Angelini, Autobahn, Boehringer-Ingelheim, Bristol-Meyers Squibb, Cardio Diagnostics, Cerevel, CNX Therapeutics, Compass Pathways, Denovo, Draig, Eli Lilly, Eumentis Therapeutics, Gedeon Richter, GH, Hikma, Holmusk, IntraCellular Therapies, Janssen/J&J, Karuna, LB Pharma, Lundbeck, MedInCell, MedLink Global, Merck, Mindpax, Maplight, Mylan, Neumora Therapeutics, Neuraxpharm, Neurocrine, Neurelis, Neurosterix, NeuShen, Neusignal Therapeutics, Newron, Noven, Novo Nordisk, Orion Pharma, Otsuka, PPD Biotech, Recognify Life Science, Recordati, Relmada, Response Pharmaetucical, Reviva, Rovi, Saladax, Sanofi, Seqirus, Servier, Sunovion, Sun Pharma, Supernus, Takeda, Teva, Terran, Tolmar, Vertex, Viatris and Xenon Pharmaceuticals. Grant/research support: Boehringer-Ingelheim, Janssen and Takeda. Honoraria/speakers or advisory boards: AbbVie, Alkermes, Allergan, Angelini, BMS, Boehringer Ingelheim, Bristol-Meyers Squibb, Cerevel, Compass, Gedeon Richter, Janssen/J&J, Karuna, LB Pharma, Lundbeck, MedInCell, Merck, Neuraxpharm, Neurelis, Neurocrine, Newron, Novo Nordisk, Otsuka, Recordati, Rovi, Seqirus, Life Science, Sunovion, Supernus, Teva, Vertex, Viatris. Stock holder/stock options: Cardio Diagnostics, Kuleon Biosciences, LB Pharmaceuticals, MedLink Global, Mindpax, Quantic, Terran. Other financial or material support: expert testimony for Janssen, Lundbeck, Neurocrine, and Otsuka; Data Safety Monitoring Board for Compass Pathways, IntraCellular Therapies, Relmada, Reviva; royalties from UpToDate and is also a stock option holder of Cardio Diagnostics, Kuleon Biosciences, LB Pharma, Medlink, Mindpax, Quantic, Terran. Fusar-Poli has received research fees from Lundbeck and honoraria from Lundbeck, Angelini, Menarini and Boehringer Ingelheim outside the current study. Mcgorry was a Founding Director of the board and now Patron of headspace (Australia) and is Executive Director of Orygen, Australia's National Centre of Excellence in Youth Mental Health. Orygen is the lead agency for five headspace centres across northwest Melbourne. He has received past unrestricted grant funding from Janssen-Cilag, AstraZeneca, Eli

Lilly, Novartis, and Pfizer, and honoraria for consultancy and teaching from Janssen-Cilag, Eli Lilly, Pfizer, AstraZeneca, Roche, Bristol Myers Squibb, and Lundbeck. He has received grant funding from the Colonial Foundation, the National Health and Medical Research Council (NHMRC) of Australia, the Australian Research Council, the National Association for Research on Schizophrenia and Depression, The Stanley Foundation, the National Institutes of Health (NIH), the Wellcome Trust, and the Australian and Victorian governments. He has been granted patents for the prevention of psychotic disorders using omega-3 polyunsaturated fatty acids in Australia and in the United States (AU 2015203289, 2017; US 9884034, 2018), with an additional claim in the United States (US 15/844444) and a pending patent application in Canada (CA 2773031). Shah received funding from salary awards from the Fonds de recherche du Quebec Sante. Vieta has received grants and served as consultant, advisor or CME speaker for the following entities: AB-Biotics, AbbVie, Adamed, Alcediag, Angelini, Biogen, Beckley-Psytech, Biohaven, Boehringer-Ingelheim, Casen-Recordati, Celon Pharma, Compass, Dainippon Sumitomo Pharma, Esteve, Ethypharm, Ferrer, Gedeon Richter, GH Research, Glaxo-Smith Kline, HMNC, Iorsia, Johnson & Johnson, Lundbeck, Luye Pharma, Medincell, Merck, Newton, Novartis, Orion Corporation, Organon, Otsuka, Roche, Rovi, Sage, Sanofi-Aventis, Sunovion, Takeda, Teva, and Viatris, outside the submitted work. Woods has received research funding from Auspex, Pfizer, and Teva, served as a consultant for Boehringer-Ingelheim and Merck, received royalties from Oxford University Press, and been granted US patent no. 8492418 B2 for a method of treating prodromal schizophrenia with glycine agonists. Solmi has received honoraria/has been a consultant for AbbVie, Angelini, Bausch Health, Boehringer Ingelheim, Lundbeck, Otsuka, Teva. Yatham reports personal fees from Canadian Network for Mood & Anxiety Treatments, Gedeon Richter, Sanofi, Intracellular Therapies, Merck, and GSK; grants and personal fees from DSP and Abbvie; and, other financial activities from CANMAT and Alkermes. Margolese reports membership on advisory board or speakers' bureaus or consultant: Abbvie, Bausch Health, BI, HLS, Janssen, Newton, Otsuka, Lundbeck, Teva; funded grants, research, or clinical trials: AlFred, Neuron, SyneuRx; Volunteer/ Advocacy: President, EPI Canada 2025-2027. Other authors have no conflict of interest to disclose. Stain: consultant for TIPS centre for clinical research in psychosis, Stavanger, Norway.

ORCID

Marco Solmi ID http://orcid.org/0000-0003-4877-7233

Gabriele Lo Buglio ID http://orcid.org/0000-0003-2425-8383

Howard C. Margolese ID http://orcid.org/0000-0002-3912-0566

Liang-Jen Wang ID http://orcid.org/0000-0002-5320-1151

Scott W. Woods ID http://orcid.org/0000-0002-3103-5228

Michael Berk ID http://orcid.org/0000-0002-5554-6946

References

NICE 2014. Psychosis and schizophrenia in adults: prevention and management: Clinical guideline.

A B, V P, S G, J G. 2010. Cognitive behavioural therapy in people at risk of psychosis. Huber.

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

35

Addington J, Addington D, Abidi S, Raedler T, Remington G. 2017. Canadian treatment guidelines for individuals at clinical high risk of psychosis. Can J Psychiatry. 62(9):656-661. https://doi.org/10.1177/0706743717719895

Addington J et al. 2011. A randomized controlled trial of cognitive behavioral therapy for individuals at clinical high risk of psychosis. Schizophr Res. 125(1):54–61. https://doi.org/10.1016/j.schres.2010.10.015

Addington J et al. 2023. Cognitive-behavioral social skills training: outcome of a randomized controlled trial for youth at risk of psychosis. Schizophr Bull Open. 4(1):sgad020. https://doi.org/10.1093/schizbullopen/sgad020

Addington J, Woods SW, Yung AR, Calkins ME, Fusar-Poli P. 2024. Harmonizing the structured interview for psychosis-risk syndromes (SIPS) and the comprehensive assessment of at-risk mental states (CAARMS): an initial approach. Early Interv Psychiatry. 18(4):248–254. https://doi.org/10.1111/eip.13401

Albert N et al. 2017. Course of illness in a sample of patients diagnosed with a schizotypal disorder and treated in a specialized early intervention setting. Findings from the 3.5-year follow-up of the OPUS II study. Schizophr Res. 182:24–30. https://doi.org/10.1016/j.schres.2016.10.013

Allen P et al. 2024. Normative modeling of brain morphometry in clinical high risk for psychosis. JAMA Psychiatry. 81(1):77. https://doi.org/10.1001/jamapsychiatry.2023.3850

Alqarni A et al. 2020. Supplementation with the omega-3 long chain polyunsaturated fatty acids: changes in the concentrations of omega-3 index, fatty acids and molecular phospholipids of people at ultra high risk of developing psychosis. Schizophr Res. 226:52–60. https://doi.org/10.1016/j.schres.2019.08.033

AMA. n.d. AMA principles of medical ethics. https://www.ama-assn.org/system/files/2019-06/code-of-medical-ethics-chapter-2.pdf

Amminger GP et al. 2020. The NEURAPRO biomarker analysis: long-chain omega-3 fatty acids improve 6-month and 12-month outcomes in youths at ultra-high risk for psychosis. Biol Psychiatry. 87(3):243-252. https://doi.org/10.1016/j.biopsych.2019.08.030

Amminger GP et al. 2010. Long-chain omega-3 fatty acids for indicated prevention of psychotic disorders: a randomized, placebo-controlled trial. Arch Gen Psychiatry. 67(2):146-154. https://doi.org/10.1001/archgenpsychiatry.2009.192

Antonucci LA et al. 2022. Using combined environmental-clinical classification models to predict role functioning outcome in clinical high-risk states for psychosis and recent-onset depression. Br J Psychiatry. 220(4):229–245. https://doi.org/10.1192/bjp.2022.16

APA. 2022. Diagnostic and statistical manual of mental disorders. 5th ed. Text Revision (DSM-5-TR).

Appiah-Kusi E et al. 2020. Effects of short-term cannabidiol treatment on response to social stress in subjects at clinical high risk of developing psychosis. Psychopharmacology. 237(4): 1121-1130. https://doi.org/10.1007/s00213-019-05442-6

Arango C et al. 2018. Preventive strategies for mental health. Lancet Psychiatry. 5(7):591-604. https://doi.org/10.1016/S2215-0366(18)30057-9

Arango C et al. 2021. Risk and protective factors for mental disorders beyond genetics: an evidence-based atlas. World Psychiatry, 20(3):417–436. https://doi.org/10.1002/wps.20894

Argolo F et al. 2024. Natural language processing in at-risk mental states: enhancing the assessment of thought disorders and psychotic traits with semantic dynamics and graph theory. Braz J Psychiatry. 46:e20233419. https://doi.org/10.47626/1516-4446-2023-3419

Arribas M et al. 2026. Joint detection of risk for psychotic disorders or bipolar disorders in clinical practice in the UK: development and validation of a clinical prediction model. Lancet Psychiatry. 13(1):14–23. https://doi.org/10.1016/S2215-0366(25)00307-4

Baldwin H et al. 2022. Neuroanatomical heterogeneity and homogeneity in individuals at clinical high risk for psychosis. Transl Psychiatry. 12(1):297. https://doi.org/10.1038/s41398-022-02057-y

Bechdolf A et al. 2023. Prevention of first-episode psychosis in people at clinical high risk: a randomized controlled, multicentre trial comparing cognitive-behavioral therapy and clinical management plus low-dose aripiprazole or placebo (PREVENT). Schizophr Bull. 49(4):1055–1066. https://doi.org/10.1093/schbul/sbad029

Bechdolf A et al. 2017. 108. PREVENT: a randomized controlled trial for the prevention of first-episode psychosis comparing cognitive-behavior therapy (CBT), clinical management, and aripiprazole combined and clinical management and placebo combined. Schizophr Bull. p. S56-S57.

Bechdolf A et al. 2011. Rationale and baseline characteristics of PREVENT: a second-generation intervention trial in subjects at-risk (prodromal) of developing first-episode psychosis evaluating cognitive behavior therapy, aripiprazole, and placebo for the prevention of psychosis. Schizophr Bull. 37(Suppl 2):S111-S121. https://doi.org/10.1093/schbul/sbr083

Bechdolf A et al. 2012. Preventing progression to first-episode psychosis in early initial prodromal states. Br J Psychiatry. 200(1):22–29. https://doi.org/10.1192/bjp.bp.109.066357

Belbasis L et al. 2018. Risk factors and peripheral biomarkers for schizophrenia spectrum disorders: an umbrella review of meta-analyses. Acta Psychiatr Scand. 137(2):88–97. https://doi.org/10.1111/acps.12847

Benrimoh D et al. 2024. On the proportion of patients who experience a prodrome prior to psychosis onset: a systematic review and meta-analysis. Mol Psychiatry. 29(5):1361-1381. https://doi.org/10.1038/s41380-024-02415-w

Berger M et al. 2019. Relationship between polyunsaturated fatty acids and psychopathology in the NEURAPRO clinical trial. Front Psychiatry. 10:393. https://doi.org/10.3389/fpsyt.2019.00393

Bhattacharyya S et al. 2018. Effect of cannabidiol on medial temporal, midbrain, and striatal dysfunction in people at clinical high risk of psychosis: a randomized clinical trial. JAMA Psychiatry. 75(11):1107-1117. https://doi.org/10.1001/jamapschiatry.2018.2309

Bolt LK et al. 2019. Neurocognition as a predictor of transition to psychotic disorder and functional outcomes in ultra-high risk participants: findings from the NEURAPRO randomized clinical trial. Schizophr Res. 206:67–74. https://doi.org/10.1016/j.schres.2018.12.013

Bortolato B et al. 2017. Systematic assessment of environmental risk factors for bipolar disorder: an umbrella review of systematic reviews and meta-analyses. Bipolar Disord. 19(2):84–96. https://doi.org/10.1111/bdi.12490

Bosnjak Kuharic D, Kekin I, Hew J, Rojnic Kuzman M, Puljak L. 2019. Interventions for prodromal stage of psychosis.

36

M. SOLMI ET AL.

Cochrane Database Syst Rev. 2019(11). https://doi.org/10.1002/14651858.CD012236.pub2

Bourgin J et al. 2020. Predicting the individual risk of psychosis conversion in at-risk mental state (ARMS): a multivariate model reveals the influence of nonpsychotic prodomal symptoms. Eur Child Adolesc Psychiatry. 29(11):1525–1535. https://doi.org/10.1007/s00787-019-01461-y

Brady LS, Larrauri CA, AMP SCZ Steering Committee. 2023. Accelerating Medicines Partnership(®) Schizophrenia (AMP(®) SCZ): developing tools to enable early intervention in the psychosis high risk state. World Psychiatry. 22(1):42-43. https://doi.org/10.1002/wps.21038

Brimblecombe N et al. 2017. The role of youth mental health services in the treatment of young people with serious mental illness: 2-year outcomes and economic implications. Early Interv Psychiatry. 11(5):393-400. https://doi.org/10.1111/eip.12261

Byrne JF et al. 2024. Proteomic biomarkers for the prediction of transition to psychosis in individuals at clinical high risk: a multi-cohort model development study. Schizophr Bull. 50(3):579–588. https://doi.org/10.1093/schbul/sbad184

Cadenhead K et al. 2017. 23. omega-3 fatty acid versus placebo in a clinical high-risk sample from the north american prodrome longitudinal studies (NAPLS) consortium. Schizophr Bull. 43(suppl_1)1:S16-S16. https://doi.org/10.1093/schbul/sbx021.042

Cadenhead KS et al. 2019. Metabolic abnormalities and low dietary Omega 3 are associated with symptom severity and worse functioning prior to the onset of psychosis: findings from the North American Prodrome Longitudinal Studies Consortium. Schizophr Res. 204:96–103. https://doi.org/10.1016/j.schres.2018.09.022

Campion J, Taylor MJ, McDaid D, Park AL, Shiers D. 2019. Applying economic models to estimate local economic benefits of improved coverage of early intervention for psychosis. Early Interv Psychiatry. 13(6):1424–1430. https://doi.org/10.1111/eip.12787

Chaumette B et al. 2016. Salivary cortisol in early psychosis: new findings and meta-analysis. Psychoneuroendocrinology. 63:262–270. https://doi.org/10.1016/j.psyneuen.2015.10.007

Choi J et al. 2017. Pupillometer-based neurofeedback cognitive training to improve processing speed and social functioning in individuals at clinical high risk for psychosis. Psychiatr Rehabil J. 40(1):33-42. https://doi.org/10.1037/prj0000217

Collin G et al. 2020. Brain functional connectivity data enhance prediction of clinical outcome in youth at risk for psychosis. Neuroimage Clin. 26:102108. https://doi.org/10.1016/j.nicl.2019.102108

Correll CU et al. 2018. Comparison of early intervention services vs treatment as usual for early-phase psychosis: a systematic review, meta-analysis, and meta-regression. JAMA Psychiatry. 75(6):555–565. https://doi.org/10.1001/jamapschiatry.2018.0623

Correll CU, Rubio JM, Kane JM. 2018. What is the risk-benefit ratio of long-term antipsychotic treatment in people with schizophrenia? World Psychiatry. 17(2):149–160. https://doi.org/10.1002/wps.20516

Correll CU et al. 2022. Mortality in people with schizophrenia: a systematic review and meta-analysis of relative risk and aggravating or attenuating factors. World Psychiatry. 21(2):248-271. https://doi.org/10.1002/wps.20994

Correll CU et al. 2017. Prevalence, incidence and mortality from cardiovascular disease in patients with pooled and specific severe mental illness: a large-scale meta-analysis of 3,211,768 patients and 113,383,368 controls. World Psychiatry. 16(2):163–180. https://doi.org/10.1002/wps.20420

Davies C et al. 2023. Hippocampal glutamate, resting perfusion and the effects of cannabidiol in psychosis risk. Schizophr Bull Open. 4(1):sgad022. https://doi.org/10.1093/schizbullopen/sgad022

Davies C et al. 2024. Increased hippocampal blood flow in people at clinical high risk for psychosis and effects of cannabidiol. Psychol Med. 54(5):993–1003. https://doi.org/10.1017/S0033291723002775

Davies C et al. 2018. Lack of evidence to favor specific preventive interventions in psychosis: a network meta-analysis. World Psychiatry. 17(2):196-209. https://doi.org/10.1002/wps.20526

Davies C et al. 2024. Connectome dysfunction in patients at clinical high risk for psychosis and modulation by oxytocin. Mol Psychiatry. 29(5):1241-1252. https://doi.org/10.1101/2023.03.17.23286528

Davies C et al. 2019. Oxytocin modulates hippocampal perfusion in people at clinical high risk for psychosis. Neuropsychopharmacology. 44(7):1300–1309. https://doi.org/10.1038/s41386-018-0311-6

Davies C et al. 2018. Efficacy and acceptability of interventions for attenuated positive psychotic symptoms in individuals at clinical high risk of psychosis: a network meta-analysis. Front Psychiatry. 9:187. https://doi.org/10.3389/fpsyt.2018.00187

Davies C et al. 2019. Neurochemical effects of oxytocin in people at clinical high risk for psychosis. Eur Neuropsychopharmacol. 29(5):601–615. https://doi.org/10.1016/j.euroneuro.2019.03.008

Davies C et al. 2020. Prenatal and perinatal risk and protective factors for psychosis: a systematic review and meta-analysis. Lancet Psychiatry. 7(5):399–410. https://doi.org/10.1016/S2215-0366(20)30057-2

Davies C et al. 2020. A single dose of cannabidiol modulates medial temporal and striatal function during fear processing in people at clinical high risk for psychosis. Transl Psychiatry. 10(1):311. https://doi.org/10.1038/s41398-020-0862-2

Davis J, Moylan S, Harvey BH, Maes M, Berk M. 2014. Neuroprogression in schizophrenia: pathways underpin-ning clinical staging and therapeutic corollaries. Aust N Z J Psychiatry. 48(6):512–529. https://doi.org/10.1177/0004867414533012

Del Fabro L et al. 2021. Functional brain network dysfunctions in subjects at high-risk for psychosis: a meta-analysis of resting-state functional connectivity. Neurosci Biobehav Rev. 128:90–101. https://doi.org/10.1016/j.neubiorev.2021.06.020

Destrée L et al. 2024. Transdiagnostic risk identification: a validation study of the Clinical High At Risk Mental State (CHARMS) criteria. Psychiatry Res. 333:115745. https://doi.org/10.1016/j.psychres.2024.115745

Devoe DJ, Farris MS, Townes P, Addington J. 2019a. Attenuated psychotic symptom interventions in youth at risk of psychosis: a systematic review and meta-analysis. Early Interv Psychiatry. 13(1):3-17. https://doi.org/10.1111/eip.12677

Devoe DJ, Farris MS, Townes P, Addington J. 2019b. Interventions and social functioning in youth at risk of psychosis: a

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

37

systematic review and meta-analysis. Early Interv Psychiatry. 13(2):169-180. https://doi.org/10.1111/eip.12689

Devoe DJ, Farris MS, Townes P, Addington J. 2020. Interventions and transition in youth at risk of psychosis: a systematic review and meta-analyses. J Clin Psychiatry. 81(3):17r12053. https://doi.org/10.4088/JCP.17r12053

Devoe DJ, Peterson A, Addington J. 2018. Negative symptom interventions in youth at risk of psychosis: a systematic review and network meta-analysis. Schizophr Bull. 44(4):807–823. https://doi.org/10.1093/schbul/sbx139

Dickens AM et al. 2021. Dysregulated lipid metabolism precedes onset of psychosis. Biol Psychiatry. 89(3):288–297. https://doi.org/10.1016/j.biopsych.2020.07.012

Ding Y et al. 2019. Brain structural abnormalities as potential markers for detecting individuals with ultra-high risk for psychosis: a systematic review and meta-analysis. Schizophr Res. 209:22–31. https://doi.org/10.1016/j.schres.2019.05.015

Doborjeh Z et al. 2023. Investigation of social and cognitive predictors in non-transition ultra-high-risk' individuals for psychosis using spiking neural networks. Schizophrenia (Heidelb). 9(1):10. https://doi.org/10.1038/s41537-023-00335-2

Dragioti E et al. 2022. Global population attributable fraction of potentially modifiable risk factors for mental disorders: a meta-umbrella systematic review. Mol Psychiatry. 27(8):3510-3519. https://doi.org/10.1038/s41380-022-01586-8

Dragioti E et al. 2019. Association of antidepressant use with adverse health outcomes: a systematic umbrella review. JAMA Psychiatry. 76(12):1241–1255. https://doi.org/10.1001/jamapschiatry.2019.2859

Dutt A et al. 2015. Exploring neural dysfunction in 'clinical high risk' for psychosis: a quantitative review of fMRI studies. J Psychiatr Res. 61:122–134. https://doi.org/10.1016/j.jpsychires.2014.08.018

Dwyer DB et al. 2022. Clinical, brain, and multilevel clustering in early psychosis and affective stages. JAMA Psychiatry. 79(7): 677-689. https://doi.org/10.1001/jamapschiatry.2022.1163

Erickson MA, Ruffle A, Gold JM. 2016. A meta-analysis of mismatch negativity in schizophrenia: from clinical risk to disease specificity and progression. Biol Psychiatry. 79(12):980–987. https://doi.org/10.1016/j.biopsych.2015.08.025

Estradé A et al. 2022. Public health primary prevention implemented by clinical high-risk services for psychosis. Transl Psychiatry. 12(1):43. https://doi.org/10.1038/s41398-022-01805-4

Firth J et al. 2020. A meta-review of "lifestyle psychiatry": the role of exercise, smoking, diet and sleep in the prevention and treatment of mental disorders. World Psychiatry. 19(3):360–380. https://doi.org/10.1002/wps.20773

French P, Morrison A. 2004. Early detection and cognitive therapy for people at high risk of developing psychosis: a treatment approach. Wiley.

Friedman-Yakoobian MS, Parrish EM, Eack SM, Keshavan MS, 2022. Neurocognitive and social cognitive training for youth at clinical high risk (CHR) for psychosis: a randomized controlled feasibility trial. Schizophr Res. 243:302–306. https://doi.org/10.1016/j.schres.2020.09.005

Fusar-Poli P et al. 2012. Predicting psychosis: meta-analysis of transition outcomes in individuals at high clinical risk. Arch Gen Psychiatry. 69(3):220–229. https://doi.org/10.1001/archgenpsychiatry.2011.1472

Fusar-Poli P et al. 2016. Towards a standard psychometric diagnostic interview for subjects at ultra high risk of

psychosis: CAARMS versus SIPS. Psychiatry J. 2016:7146341-7146311. https://doi.org/10.1155/2016/7146341

Fusar-Poli P et al. 2021. Preventive psychiatry: a blueprint for improving the mental health of young people. World Psychiatry. 20(2):200-221. https://doi.org/10.1002/wps.20869

Fusar-Poli P et al. 2019. Preventive treatments for psychosis: umbrella review (just the evidence). Front Psychiatry. 10:764. https://doi.org/10.3389/fpsyt.2019.00764

Fusar-Poli P et al. 2020. Real-world clinical outcomes two years after transition to psychosis in individuals at clinical high risk: electronic health record cohort study. Schizophr Bull. 46(5):1114–1125. https://doi.org/10.1093/schbul/sbaa040

Fusar-Poli P, Hijazi Z, Stahl D, Steyerberg EW. 2018. The science of prognosis in psychiatry: a review. JAMA Psychiatry. 75(12):1289-1297. https://doi.org/10.1001/jamapsychiatry.2018.2530

Fusar-Poli P et al. 2022. Ethical considerations for precision psychiatry: a roadmap for research and clinical practice. Eur Neuropsychopharmacol. 63:17–34. https://doi.org/10.1016/j.euroneuro.2022.08.001

Fusar-Poli P, Radua J. 2018. Ten simple rules for conducting umbrella reviews. Evid Based Ment Health. 21(3):95-100. https://doi.org/10.1136/ebmental-2018-300014

Fusar-Poli P, Radua J, Davies C, Jauhar S. 2022. Overoptimistic literature and methodological biases favoring cognitive behavioral therapy for the prevention of psychosis. Schizophr Bull. 48(1):1-3. https://doi.org/10.1093/schbul/sbab129

Fusar-Poli P, Radua J, Jauhar S. 2021. Lack of robust meta-analytic evidence to favour cognitive behavioural therapy for prevention of psychosis. World Psychiatry. 20(3):443-444. https://doi.org/10.1002/wps.20896

Fusar-Poli P et al. 2017. Development and validation of a clinically based risk calculator for the transdiagnostic prediction of psychosis. JAMA Psychiatry. 74(5):493-500. https://doi.org/10.1001/jamapsychiatry.2017.0284

Fusar-Poli P et al. 2017. Long-term validity of the At Risk Mental State (ARMS) for predicting psychotic and non-psychotic mental disorders. Eur Psychiatry. 42:49–54. https://doi.org/10.1016/j.eurpsy.2016.11.010

Fusar-Poli P et al. 2020. Prevention of psychosis: advances in detection, prognosis, and intervention. JAMA Psychiatry. 77(7): 755-765. https://doi.org/10.1001/jamapschiatry.2019.4779

Fusar-Poli P et al. 2016. The dark side of the moon: meta-analytical impact of recruitment strategies on risk enrichment in the clinical high risk state for psychosis. Schizophr Bull. 42(3):732-743. https://doi.org/10.1093/schbul/sbv162

Fusar-Poli P et al. 2019. Transdiagnostic psychiatry: a systematic review. World Psychiatry. 18(2):192-207. https://doi.org/10.1002/wps.20631

Fusar-Poli P et al. 2019. Clinical-learning versus machine-learning for transdiagnostic prediction of psychosis onset in individuals at-risk. Transl Psychiatry. 9(1):259. https://doi.org/10.1038/s41398-019-0600-9

Fusar-Poli P, Sullivan SA, Shah JL, Uhlhaas PJ. 2019. Improving the detection of individuals at clinical risk for psychosis in the community, primary and secondary care: an integrated evidence-based approach. Front Psychiatry. 10:774. https://doi.org/10.3389/fpsyt.2019.00774

Fusar-Poli P et al. 2019. Transdiagnostic risk calculator for the automatic detection of individuals at risk and the

38

M. SOLMI ET AL.

prediction of psychosis: second replication in an independent national health service trust. Schizophr Bull. 45(3):562-570. https://doi.org/10.1093/schbul/sby070

Gao X et al. 2022. Unsaturated fatty acids in mental disorders: an umbrella review of meta-analyses. Adv Nutr. 13(6):2217-2236. https://doi.org/10.1093/advances/nmac084

Gates M et al. 2022. Reporting guideline for overviews of reviews of healthcare interventions: development of the PRIOR statement. BMJ. 378:e070849. https://doi.org/10.1136/bmj-2022-070849

Gerstenberg M et al. 2015. Frequency and correlates of DSM-5 attenuated psychosis syndrome in a sample of adolescent inpatients with nonpsychotic psychiatric disorders. J Clin Psychiatry. 76(11):e1449-1458-e1458. https://doi.org/10.4088/JCP.14m09435

Gerstenberg M et al. 2016. Adolescents and adults at clinical high-risk for psychosis: age-related differences in attenuated positive symptoms syndrome prevalence and entanglement with basic symptoms. Psychol Med. 46(5):1069–1078. https://doi.org/10.1017/S0033291715002627

Glenthøj LB, Kristensen TD, Wenneberg C, Hjorthøj C, Nordentoft M. 2021. Predictors of remission from the ultra-high risk state for psychosis. Early Interv Psychiatry. 15(1):104–112. https://doi.org/10.1111/eip.12917

Glenthøj LB et al. 2020. Cognitive remediation plus standard treatment versus standard treatment alone for individuals at ultra-high risk of developing psychosis: results of the FOCUS randomised clinical trial. Schizophr Res. 224:151–158. https://doi.org/10.1016/j.schres.2020.08.016

Goodman S, Hammond D. 2022. Perceptions of the health risks of cannabis: estimates from national surveys in Canada and the United States, 2018-2019. Health Educ Res. 37(2):61-78. https://doi.org/10.1093/her/cyac006

Grent-'t-Jong T et al. 2021. 40-Hz auditory steady-state responses characterize circuit dysfunctions and predict clinical outcomes in clinical high-risk for psychosis participants: a magnetoencephalography study. Biol Psychiatry. 90(6):419–429. https://doi.org/10.1016/j.biopsych.2021.03.018

Gurusamy J, Gandhi S, Damodharan D, Ganesan V, Palaniappan M. 2018. Exercise, diet and educational interventions for metabolic syndrome in persons with schizophrenia: a systematic review. Asian J Psychiatr. 36:73–85. https://doi.org/10.1016/j.ajp.2018.06.018

Haining K et al. 2021. The relationship between cognitive deficits and impaired short-term functional outcome in clinical high-risk for psychosis participants: a machine learning and modelling approach. Schizophr Res. 231:24–31. https://doi.org/10.1016/j.schres.2021.02.019

Harrisberger F et al. 2016. Alterations in the hippocampus and thalamus in individuals at high risk for psychosis. NPJ Schizophr. 2(1):16033. https://doi.org/10.1038/npjschz.2016.33

Hart XM et al. 2024. Optimisation of pharmacotherapy in psychiatry through therapeutic drug monitoring, molecular brain imaging and pharmacogenetic tests: focus on antipsychotics. World J Biol Psychiatry. 25(9):451-536. https://doi.org/10.1080/15622975.2024.2366235

Hartmann S et al. 2024. Development and temporal validation of a clinical prediction model of transition to psychosis in individuals at ultra-high risk in the UHR 1000+ cohort. World Psychiatry. 23(3):400-410. https://doi.org/10.1002/wps.21240

Hasan A et al. 2019. WFSBP guidelines on how to grade treatment evidence for clinical guideline development. World J Biol Psychiatry. 20(1):2–16. https://doi.org/10.1080/15622975.2018.1557346

Hauke DJ et al. 2021. Multimodal prognosis of negative symptom severity in individuals at increased risk of developing psychosis. Transl Psychiatry. 11(1):312. https://doi.org/10.1038/s41398-021-01409-4

Hogg B et al. 2023. Psychological trauma as a transdiagnostic risk factor for mental disorder: an umbrella meta-analysis. Eur Arch Psychiatry Clin Neurosci. 273(2):397–410. https://doi.org/10.1007/s00406-022-01495-5

Howes OD et al. 2021. The clinical significance of duration of untreated psychosis: an umbrella review and random-effects meta-analysis. World Psychiatry. 20(1):75–95. https://doi.org/10.1002/wps.20822

Hutton P, Taylor PJ. 2014. Cognitive behavioural therapy for psychosis prevention: a systematic review and meta-analysis. Psychol Med. 44(3):449-468. https://doi.org/10.1017/S0033291713000354

Iftimovici A, Group IC-pS. et al. 2023. Asynchronous neural maturation predicts transition to psychosis. Psychiatry Clin Neurosci. https://doi.org/10.1111/pcn.13612

loannidis JP. 2009. Integration of evidence from multiple meta-analyses: a primer on umbrella reviews, treatment networks and multiple treatments meta-analyses. CMAJ. 181(8):488–493. https://doi.org/10.1503/cmaj.081086

Irving J et al. 2021. Using natural language processing on electronic health records to enhance detection and prediction of psychosis risk. Schizophr Bull. 47(2):405–414. https://doi.org/10.1093/schbul/sbaa126

Ising HK et al. 2016. Four-year follow-up of cognitive behavioral therapy in persons at ultra-high risk for developing psychosis: the dutch early detection intervention evaluation (EDIE-NL) Trial. Schizophr Bull. 42(5):1243-1252. https://doi.org/10.1093/schbul/sbw018

Ising HK et al. 2017. Four-year cost-effectiveness of cognitive behavior therapy for preventing first-episode psychosis: the dutch early detection intervention evaluation (EDIE-NL) trial. Schizophr Bull. 43(2):365-374. https://doi.org/10.1093/schbul/sbw084

Ising HK et al. 2015. Cost-effectiveness of preventing first-episode psychosis in ultra-high-risk subjects: multi-centre randomized controlled trial. Psychol Med. 45(7):1435–1446. https://doi.org/10.1017/S0033291714002530

Jalbrzikowski M et al. 2021. Association of structural magnetic resonance imaging measures with psychosis onset in individuals at clinical high risk for developing psychosis: an ENIGMA working group mega-analysis. JAMA Psychiatry. 78(7):753-766. https://doi.org/10.1001/jamapsychiatry.2021.0638

Jin H, Tappenden P, MacCabe JH, Robinson S, Byford S. 2020. Evaluation of the cost-effectiveness of services for schizophrenia in the UK across the entire care pathway in a single whole-disease model. JAMA Netw Open. 3(5):e205888. https://doi.org/10.1001/jamanetworkopen.2020.5888

Jin H et al. 2021. Cost and health impacts of adherence to the national institute for health and care excellence schizophrenia guideline recommendations. Br J Psychiatry. 218(4):224–229. https://doi.org/10.1192/bjp.2020.241

Kahn RS et al. 2015. Schizophrenia. Nat Rev Dis Primers. 1(1):15067. https://doi.org/10.1038/nrdp.2015.67

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

39

Kantrowitz JT et al. 2015. D-serine for the treatment of negative symptoms in individuals at clinical high risk of schizophrenia: a pilot, double-blind, placebo-controlled, randomised parallel group mechanistic proof-of-concept trial. Lancet Psychiatry. 2(5):403–412. https://doi.org/10.1016/S2215-0366(15)00098-X

Kelleher I et al. 2012. Prevalence of psychotic symptoms in childhood and adolescence: a systematic review and meta-analysis of population-based studies. Psychol Med. 42(9): 1857–1863. https://doi.org/10.1017/S0033291711002960

Kelleher I et al. 2012. Identification and characterization of prodromal risk syndromes in young adolescents in the community: a population-based clinical interview study. Schizophr Bull. 38(2):239-246. https://doi.org/10.1093/schbul/sbr164

Kim JY et al. 2019. Environmental risk factors and biomarkers for autism spectrum disorder: an umbrella review of the evidence. Lancet Psychiatry. 6(7):590–600. https://doi.org/10.1016/S2215-0366(19)30181-6

Kizilay E et al. 2024. Automated linguistic analysis in youth at clinical high risk for psychosis. Schizophr Res. 274:121-128. https://doi.org/10.1016/j.schres.2024.09.009

Kotlicka-Antczak M et al. 2020. Worldwide implementation of clinical services for the prevention of psychosis: the IEPA early intervention in mental health survey. Early Interv Psychiatry. 14(6):741-750. https://doi.org/10.1111/eip.12950

Koutsouleris N et al. 2021. Multimodal machine learning workflows for prediction of psychosis in patients with clinical high-risk syndromes and recent-onset depression. JAMA Psychiatry. 78(2):195-209. https://doi.org/10.1001/jamapsychiatry.2020.3604

Koutsouleris N et al. 2022. Exploring links between psychosis and frontotemporal dementia using multimodal machine learning: dementia praecox revisited. JAMA Psychiatry. 79(9): 907–919. https://doi.org/10.1001/jamapsychiatry.2022.2075

Krakowski K, Oliver D, Arribas M, Stahl D, Fusar-Poli P. 2024. Dynamic and transdiagnostic risk calculator based on natural language processing for the prediction of psychosis in secondary mental health care: development and internal-external validation cohort study. Biol Psychiatry. 96(7):604-614. https://doi.org/10.1016/j.biopsych.2024.05.022

Kristensen TD et al. 2020. No effects of cognitive remediation on cerebral white matter in individuals at ultra-high risk for psychosis-A randomized clinical trial. Front Psychiatry. 11:873. https://doi.org/10.3389/fpsyt.2020.00873

Lång U et al. 2022. Systematic review and meta-analysis: psychosis risk in children and adolescents with an at-risk mental state. J Am Acad Child Adolesc Psychiatry. 61(5): 615-625. https://doi.org/10.1016/j.jaac.2021.07.593

Lee TY et al. 2022. Prediction of psychosis: model develop-ment and internal validation of a personalized risk calcula-tor. Psychol Med. 52(13):2632-2640. https://doi.org/10.1017/S0033291720004675

Li H et al. 2021. Enhancing attention and memory of individuals at clinical high risk for psychosis with mHealth technology. Asian J Psychiatr. 58:102587. https://doi.org/10.1016/j.ajp.2021.102587

Li Z et al. 2022. Plasma metabolic alterations and potential biomarkers in individuals at clinical high risk for psychosis. Schizophr Res. 239:19–28. https://doi.org/10.1016/j.schres.2021.11.011

Liu CC et al. 2015. Neurocognitive functioning of subjects with putative pre-psychotic states and early psychosis.

Schizophr Res. 164(1-3):40–46. https://doi.org/10.1016/j.schres.2015.03.006

Livingston NR et al. 2024. Effects of diazepam on hippocampal blood flow in people at clinical high risk for psychosis. Neuropsychopharmacology. 49(9):1448–1458. https://doi.org/10.1038/s41386-024-01864-9

Loch AA et al. 2023. Detecting at-risk mental states for psychosis (ARMS) using machine learning ensembles and facial features. Schizophr Res. 258:45–52. https://doi.org/10.1016/j.schres.2023.07.011

Loewy R et al. 2016. Intensive auditory cognitive training improves verbal memory in adolescents and young adults at clinical high risk for psychosis. Schizophr Bull. 42 Suppl 1(Suppl 1):S118-S126. https://doi.org/10.1093/schbul/sbw009

Luna LP et al. 2022. A systematic review and meta-analysis of structural and functional brain alterations in individuals with genetic and clinical high-risk for psychosis and bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry. 117:110540. https://doi.org/10.1016/i.pnpbp.2022.110540

Lyu H et al. 2023. Eye movement abnormalities can distinguish first-episode schizophrenia, chronic schizophrenia, and prodromal patients from healthy controls. Schizophr Bull Open. 4(1):sgac076. https://doi.org/10.1093/schizbullopen/sgac076

Mallawaarachchi SR et al. 2020. Cognitive functioning in ultra-high risk for psychosis individuals with and without depression: secondary analysis of findings from the NEURAPRO randomized clinical trial. Schizophr Res. 218:48–54. https://doi.org/10.1016/j.schres.2020.03.008

McCutcheon RA et al. 2025. INTEGRATE: international guidelines for the algorithmic treatment of schizophrenia. Lancet Psychiatry. 12(5):384–394. https://doi.org/10.1016/S2215-0366(25)00031-8

McGlashan TH et al. 2006. Randomized, double-blind trial of olanzapine versus placebo in patients prodomally symptomatic for psychosis. Am J Psychiatry. 163(5):790-799. https://doi.org/10.1176/ajp.2006.163.5.790

McGorry PD, Hartmann JA, Spooner R, Nelson B. 2018. Beyond the "at risk mental state" concept: transitioning to transdiagnostic psychiatry. World Psychiatry. 17(2):133–142. https://doi.org/10.1002/wps.20514

McGorry PD et al. 2023. A Sequential adaptive intervention strategy targeting remission and functional recovery in young people at ultrahigh risk of psychosis: the staged treatment in early psychosis (STEP) sequential multiple assignment randomized trial. JAMA Psychiatry. 80(9):875–885. https://doi.org/10.1001/jamappsychiatry.2023.1947

McGorry PD et al. 2022. Designing and scaling up integrated youth mental health care. World Psychiatry. 21(1):61-76. https://doi.org/10.1002/wps.20938

McGorry PD et al. 2024. The lancet psychiatry commission on youth mental health. Lancet Psychiatry. 11(9):731-774. https://doi.org/10.1016/S2215-0366(24)00163-9

McGorry PD et al. 2017. Effect of ω-3 polyunsaturated fatty acids in young people at ultrahigh risk for psychotic disorders: the NEURAPRO randomized clinical trial. JAMA Psychiatry. 74(1):19–27. https://doi.org/10.1001/jamapschiatry.2016.2902

McGorry PD et al. 2013. Randomized controlled trial of interventions for young people at ultra-high risk of psychosis: twelve-month outcome. J Clin Psychiatry. 74(4):349-356. https://doi.org/10.4088/JCP.12m07785

40

M. SOLMI ET AL.

McGorry PD et al. 2002. Randomized controlled trial of interventions designed to reduce the risk of progression to first-episode psychosis in a clinical sample with subthreshold symptoms. Arch Gen Psychiatry. 59(10):921-928. https://doi.org/10.1001/archpsyc.59.10.921

Meehan AJ et al. 2022. Clinical prediction models in psychiatry: a systematic review of two decades of progress and challenges. Mol Psychiatry. 27(6):2700-2708. https://doi.org/10.1038/s41380-022-01528-4

Mei C et al. 2021. Preventive interventions for individuals at ultra high risk for psychosis: an updated and extended meta-analysis. Clin Psychol Rev. 86:102005. https://doi.org/10.1016/j.cpr.2021.102005

Merritt K, Egerton A, Kempton MJ, Taylor MJ, McGuire PK. 2016. Nature of glutamate alterations in schizophrenia: a meta-analysis of proton magnetic resonance spectroscopy studies. JAMA Psychiatry. 73(7):665–674. https://doi.org/10.1001/jamapsychiatry.2016.0442

Miklowitz DJ et al. 2014. Family-focused treatment for adolescents and young adults at high risk for psychosis: results of a randomized trial. J Am Acad Child Adolesc Psychiatry. 53(8):848-858. https://doi.org/10.1016/j.jaac.2014.04.020

Miller TJ et al. 2003. Prodromal assessment with the structured interview for prodromal syndromes and the scale of prodromal symptoms: predictive validity, interrater reliability, and training to reliability. Schizophr Bull. 29(4):703-715. https://doi.org/10.1093/oxfordjournals.schbul.a007040

Minichino A et al. 2025. Preventing psychosis in people at clinical high risk: an updated meta-analysis by the world psychiatric association preventive psychiatry section. Mol Psychiatry. 30(6):2773-2782. https://doi.org/10.1038/s41380-025-02902-8

Mondelli V et al. 2023. Serum immune markers and transition to psychosis in individuals at clinical high risk. Brain Behav Immun. 110:290–296. https://doi.org/10.1016/j.bbi.2023.03.014

Mongan D et al. 2021. Development of proteomic prediction models for transition to psychotic disorder in the clinical high-risk state and psychotic experiences in adolescence. JAMA Psychiatry. 78(1):77–90. https://doi.org/10.1001/jampsychiatry.2020.2459

Moore TM et al. 2022. Development of a probability calculator for psychosis risk in children, adolescents, and young adults. Psychol Med. 52(14):3159-3167. https://doi.org/10.1017/S0033291720005231

Morrison AP et al. 2007. Three-year follow-up of a randomized controlled trial of cognitive therapy for the prevention of psychosis in people at ultrahigh risk. Schizophr Bull. 33(3):682-687. https://doi.org/10.1093/schbul/sbl042

Morrison AP et al. 2012. Early detection and intervention evaluation for people at risk of psychosis: multisite randomised controlled trial. BMJ. 344(apr05 1):e2233-e2233. https://doi.org/10.1136/bmj.e2233

Morrison AP et al. 2004. Cognitive therapy for the prevention of psychosis in people at ultra-high risk: randomised controlled trial. Br J Psychiatry. 185(4):291–297. https://doi.org/10.1192/bjp.185.4.291

Mossaheb N et al. 2018. Changes in triglyceride levels in ultra-high risk for psychosis individuals treated with omega-3 fatty acids. Early Interv Psychiatry. 12(1):30–36. https://doi.org/10.1111/eip.12275

Mossaheb N et al. 2018. Predictors of longer-term outcome in the Vienna omega-3 high-risk study. Schizophr Res. 193:168–172. https://doi.org/10.1016/j.schres.2017.08.010

Myin-Germeys I et al. 2021. Efficacy of Acceptance and Commitment Therapy in Daily Life in Early Psychosis: Results from the Multi-Center INTERACT Randomized Controlled Trial. Psychother Psychosom. 2022;91(6):411-423. https://doi.org/10.1159/000522274

Myran DT et al. 2023. Transition to schizophrenia spectrum disorder following emergency department visits due to substance use with and without psychosis. JAMA Psychiatry. 80(11):1169–1174. https://doi.org/10.1001/jamappsychiatry.2023.3582

Myran DT et al. 2024. Development of an anxiety disorder following an emergency department visit due to cannabis use: a population-based cohort study. EClinicalMedicine. 69:102455. https://doi.org/10.1016/j.eclinm.2024.102455

Myran DT et al. 2025. Changes in incident schizophrenia diagnoses associated with cannabis use disorder after cannabis legalization. JAMA Netw Open. 8(2):e2457868. https://doi.org/10.1001/jamanetworkopen.2024.57868

Myran DT et al. 2025. Cannabis use disorder emergency department visits and hospitalizations and 5-year mortality. JAMA Netw Open. 8(2):e2457852. https://doi.org/10.1001/jamanetworkopen.2024.57852

Myran DT et al. 2023. Association between non-medical cannabis legalization and emergency department visits for cannabis-induced psychosis. Mol Psychiatry. 28(10):4251-4260. https://doi.org/10.1038/s41380-023-02185-x

Nelson B et al. 2018. NEURAPRO: a multi-centre RCT of omega-3 polyunsaturated fatty acids versus placebo in young people at ultra-high risk of psychotic disorders-medium-term follow-up and clinical course. NPJ Schizophr. 4(1):11. https://doi.org/10.1038/s41537-018-0052-x

Nordentoft M et al. 2006. Transition rates from schizotypal disorder to psychotic disorder for first-contact patients included in the OPUS trial. A randomized clinical trial of integrated treatment and standard treatment. Schizophr Res. 83(1):29–40. https://doi.org/10.1016/j.schres. 2006.01.002

Oliver D et al. 2022. Prognostic accuracy and clinical utility of psychometric instruments for individuals at clinical high-risk of psychosis: a systematic review and meta-analysis. Mol Psychiatry. 27(9):3670-3678. https://doi.org/10.1038/s41380-022-01611-w

Oliver D et al. 2018. Meta-analytical prognostic accuracy of the Comprehensive Assessment of at Risk Mental States (CAARMS): the need for refined prediction. Eur Psychiatry. 49:62–68. https://doi.org/10.1016/j.eurpsy.2017.10.001

Oliver D et al. 2020. What causes the onset of psychosis in individuals at clinical high risk? a meta-analysis of risk and protective factors. Schizophr Bull. 46(1):110–120. https://doi.org/10.1093/schbul/sbz039

Oliver D et al. 2020. Transdiagnostic individualized clinically-based risk calculator for the automatic detection of individuals at-risk and the prediction of psychosis: external replication in 2,430,333 US patients. Transl Psychiatry. 10(1):364. https://doi.org/10.1038/s41398-020-01032-9

Ologundudu OM et al. 2023. Risk stratification for treating people at ultra-high risk for psychosis: a cost-effectiveness analysis. Schizophr Res. 261:225-233. https://doi.org/10.1016/j.schres.2023.09.015

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

41

Osborne KJ, Mittal VA. 2019. External validation and extension of the NAPLS-2 and SIPS-RC personalized risk calculators in an independent clinical high-risk sample. Psychiatry Res. 279:9-14. https://doi.org/10.1016/j.psychres.2019.06.034

Page MJ et al. 2021. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Rev Esp Cardiol (Engl Ed). 74(9):790-799. https://doi.org/10.1016/j.rec.2021.07.010

Park S, Miller BJ. 2020. Meta-analysis of cytokine and C-reactive protein levels in high-risk psychosis. Schizophr Res. 226:5–12. https://doi.org/10.1016/j.schres.2019.03.012

Perez J et al. 2015. Clinical effectiveness and cost-effectiveness of tailored intensive liaison between primary and secondary care to identify individuals at risk of a first psychotic illness (the LEGs study): a cluster-randomised controlled trial. Lancet Psychiatry. 2(11):984–993. https://doi.org/10.1016/s2215-0366(15)00157-1

Perkins DO et al. 2020. Polygenic risk score contribution to psychosis prediction in a target population of persons at clinical high risk. Am J Psychiatry. 177(2):155–163. https://doi.org/10.1176/appi.ajp.2019.18060721

Petersen L et al. 2005. A randomised multicentre trial of integrated versus standard treatment for patients with a first episode of psychotic illness. BMJ. 331(7517):602. https://doi.org/10.1136/bmj.38565.415000.E01

Phillips LJ et al. 2009. Cost implications of specific and non-specific treatment for young persons at ultra high risk of developing a first episode of psychosis. Early Interv Psychiatry. 3(1):28-34. https://doi.org/10.1111/j.1751-7893. 2008.00106.x

Piskulic D, Barbato M, Liu L, Addington J. 2015. Pilot study of cognitive remediation therapy on cognition in young people at clinical high risk of psychosis. Psychiatry Res. 225(1-2):93-98. https://doi.org/10.1016/j.psychres.2014.10.021

Piskulic D, Romanowska S, Addington J. 2018. Pilot study of cognitive remediation and motivational interviewing in youth at risk of serious mental illness. Early Interv Psychiatry. 12(6):1193–1197. https://doi.org/10.1111/eip.12520

Pozza A, Domenichetti S, Dèttore D. 2019. Terapia cognitivo-comportamentale per giovani ad alto rischio di un primo episodio di psicosi: un trial randomizzato controllato. Riv Psichiatr. 54(4):150–159. https://doi.org/10.1708/3202.31797

Pozza A, Domenichetti S, Dèttore D. 2021. Cognitive behavioural therapy for help-seeking adolescents and young adults with at-risk-mental state: effects on subclinical positive symptoms. Early Interv Psychiatry. 15(3):513–524. https://doi.org/10.1111/eip.12974

Preti A, Cella M. 2010. Randomized-controlled trials in people at ultra high risk of psychosis: a review of treatment effectiveness. Schizophr Res. 123(1):30–36. https://doi.org/10.1016/j.schres.2010.07.026

Puntis S, Oliver D, Fusar-Poli P. 2021. Third external replication of an individualised transdiagnostic prediction model for the automatic detection of individuals at risk of psychosis using electronic health records. Schizophr Res. 228:403–409. https://doi.org/10.1016/j.schres.2021.01.005

Qurashi I et al. 2024. A randomised double-blind placebo-controlled trial of minocycline and/or omega-3 fatty acids added to treatment as usual for at risk Mental States: the NAYAB study. Brain Behav Immun. 115:609–616. https://doi.org/10.1016/j.bbi.2023.10.025

Radua J et al. 2018. What causes psychosis? an umbrella review of risk and protective factors. World Psychiatry. 17(1):49–66. https://doi.org/10.1002/wps.20490

Raket LL et al. 2020. Dynamic ElecTronic hEalth reCord de- Tection (DETECT) of individuals at risk of a first episode of psychosis: a case-control development and validation study. Lancet Digit Health. 2(5):e229-e239. https://doi. org/10.1016/S2589-7500(20)30024-8

Rambaud V, Marzo A, Chaumette B. 2022. Oxidative stress and emergence of psychosis. Antioxidants (Basel). 11(10):1870. https://doi.org/10.3390/antiox11101870

Reinen JM et al. 2024. Multimodal fusion of brain signals for robust prediction of psychosis transition. Schizophrenia (Heidelb). 10(1):54. https://doi.org/10.1038/s41537-024-00464-2

Reininghaus U et al. 2019. Efficacy of acceptance and commitment therapy in daily life (ACT-DL) in early psychosis: study protocol for a multi-centre randomized controlled trial. Trials. 20(1):769. https://doi.org/10.1186/s13063-019-3912-4

Rinne GR, O'Brien MP, Miklowitz DJ, Addington JM, Cannon TD. 2021. Depression, family interaction and family intervention in adolescents at clinical-high risk for psychosis. Early Interv Psychiatry. 15(2):360–366. https://doi.org/10.1111/eip.12954

Rosson S et al. 2022. Brain stimulation and other biological non-pharmacological interventions in mental disorders: an umbrella review. Neurosci Biobehav Rev. 139:104743. https://doi.org/10.1016/j.neubiorev.2022.104743

Ruhrmann S et al. 2007. Acute effects of treatment for pro-dromal symptoms for people putatively in a late initial prodromal state of psychosis. Br J Psychiatry Suppl. 51:S88-S95. https://doi.org/10.1192/bjp.191.51.s88

Salazar de Pablo G et al. 2021. Transition to psychosis in randomized clinical trials of individuals at clinical high risk of psychosis compared to observational cohorts: a systematic review and meta-analysis. Eur Psychiatry. 64(1):e51. https://doi.org/10.1192/j.eurpsy.2021.2222

Salazar de Pablo G et al. 2026. Development and validation of a prognostic model and risk calculator for the estimation of bipolar-spectrum disorder risk in hospitalised adolescents with non-psychotic/non-bipolar mental disorders. Mol Psychiatry. 31(3):1333–1342. https://doi.org/10.1038/s41380-025-03244-1

Salazar de Pablo G et al. 2021. Probability of transition to psychosis in individuals at clinical high risk: an updated meta-analysis. JAMA Psychiatry. 78(9):970–978. https://doi.org/10.1001/jamapsychiatry.2021.0830

Salazar de Pablo G et al. 2021. Implementing precision psychiatry: a systematic review of individualized prediction models for clinical practice. Schizophr Bull. 47(2):284–297. https://doi.org/10.1093/schbul/sbaa120

Salazar de Pablo G, Woods SW, Drymonitou G, de Diego H, Fusar-Poli P. 2021. Prevalence of individuals at clinical high-risk of psychosis in the general population and clinical samples: systematic review and meta-analysis. Brain Sci. 11(11):1544. https://doi.org/10.3390/brainsci11111544

Schimmelmann BG, Michel C, Martz-lrngartinger A, Linder C, Schultze-Lutter F. 2015. Age matters in the prevalence and clinical significance of ultra-high-risk for psychosis symptoms and criteria in the general population: findings from

42

M. SOLMI ET AL.

the BEAR and BEARS-kid studies. World Psychiatry. 14(2):189-197. https://doi.org/10.1002/wps.20216

Schmidt A et al. 2020. Acute oxytocin effects in inferring others' beliefs and social emotions in people at clinical high risk for psychosis. Transl Psychiatry. 10(1):203. https://doi.org/10.1038/s41398-020-00885-4

Schmidt SJ et al. 2015. EPA guidance on the early interven-tion in clinical high risk states of psychoses. Eur Psychiatry. 30(3):388-404. https://doi.org/10.1016/j.eurpsy.2015.01.013

Schoretsanitis G et al. 2024. The European psychiatric association (EPA) - early career psychiatrists committee survey on trainees' and early-career psychiatrists' attitudes towards therapeutic drug monitoring (TDM) use and utility during antipsychotic treatment. World J Biol Psychiatry. 25(6):342–351. https://doi.org/10.1080/15622975.2024.2367138

Schultze-Lutter F et al. 2012. Basic symptoms and the predic-tion of first-episode psychosis. Curr Pharm Des. 18(4):351-357. https://doi.org/10.2174/138161212799316064

Shah JL et al. 2017. Is the clinical high-risk state a valid concept? retrospective examination in a first-episode psychosis sample. Psychiatr Serv. 68(10):1046–1052. https://doi.org/10.1176/appi.ps.201600304

Sherman BJ, McRae-Clark AL. 2016. Treatment of cannabis use disorder: current science and future outlook. Pharmacotherapy. 36(5):511-535. https://doi.org/10.1002/phar.1747

Simmonite M, Steeby CJ, Taylor SF. 2023. Medial frontal cortex gaba concentrations in psychosis spectrum and mood disorders: a meta-analysis of proton magnetic resonance spectroscopy studies. Biol Psychiatry. 93(2):125–136. https://doi.org/10.1016/j.biopsych.2022.08.004

Sizer H et al. 2022. Outcomes for first-episode psychosis after entry via an at-risk mental state clinic compared to direct entry to a first episode of psychosis service: a systematic review and meta-analysis. Schizophr Res. 240:214–219. https://doi.org/10.1016/j.schres.2021.12.019

Solmi M et al. 2021. Risk and protective factors for alcohol and tobacco related disorders: an umbrella review of observational studies. Neurosci Biobehav Rev. 121:20–28. https://doi.org/10.1016/j.neubiorev.2020.11.010

Solmi M et al. 2023. Efficacy and acceptability of psychosocial interventions in schizophrenia: systematic overview and quality appraisal of the meta-analytic evidence. Mol. Psychiatry. 28(1):354-368. https://doi.org/10.1038/s41380-022-01727-z

Solmi M et al. 2023. Balancing risks and benefits of cannabis use: umbrella review of meta-analyses of randomised controlled trials and observational studies. BMJ. 382:e072348. https://doi.org/10.1136/bmj-2022-072348

Solmi M et al. 2021. Risk and protective factors for mental disorders with onset in childhood/adolescence: an umbrella review of published meta-analyses of observational longitudinal studies. Neurosci Biobehav Rev. 120:565–573. https://doi.org/10.1016/j.neubiorev.2020.09.002

Solmi M et al. 2021. Risk and protective factors for personality disorders: an umbrella review of published meta-analyses of case-control and cohort studies. Front Psychiatry. 12:679379. https://doi.org/10.3389/fpsyt.2021.679379

Solmi M et al. 2021. Risk and protective factors for cannabis, cocaine, and opioid use disorders: an umbrella review of meta-analyses of observational studies. Neurosci

Biobehav Rev. 126:243-251. https://doi.org/10.1016/j.neubiorev.2021.03.014

Solmi M et al. 2021. Disparities in screening and treatment of cardiovascular diseases in patients with mental disorders across the world: systematic review and meta-analysis of 47 observational studies. Am J Psychiatry. 178(9):793-803. https://doi.org/10.1176/appi.ajp.2021.21010031

Solmi M et al. 2020. Disparities in cancer screening in people with mental illness across the world versus the general population: prevalence and comparative meta-analysis including 4717 839 people. Lancet Psychiatry. 7(1):52-63. https://doi.org/10.1016/S2215-0366(19)30414-6

Solmi M et al. 2017. Safety, tolerability, and risks associated with first- and second-generation antipsychotics: a state-of-the-art clinical review. Ther Clin Risk Manag. 13:757-777. https://doi.org/10.2147/TCRM.S117321

Solmi M et al. 2022. Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies. Mol Psychiatry. 27(1):281-295. https://doi.org/10.1038/s41380-021-01161-7

Solmi M et al. 2021. Risk factors for eating disorders: an umbrella review of published meta-analyses. Braz J Psychiatry. 43(3):314–323. https://doi.org/10.1590/1516-4446-2020-1099

Solmi M et al. 2023. Incidence, prevalence, and global burden of schizophrenia - data, with critical appraisal, from the Global Burden of Disease (GBD) 2019. Mol Psychiatry. 28(12): 5319–5327. https://doi.org/10.1038/s41380-023-02138-4

Solmi M et al. 2023. Meta-analytic prevalence of comorbid mental disorders in individuals at clinical high risk of psychosis: the case for transdiagnostic assessment. Mol Psychiatry. 28(6):2291-2300. https://doi.org/10.1038/s41380-023-02029-8

Solmi M et al. 2020. Factors associated with loneliness: an umbrella review of observational studies. J Affect Disord. 271:131–138. https://doi.org/10.1016/j.jad.2020.03.075

Sprungli-Toffel E et al. 2024. Individualized pretest risk estimates to guide treatment decisions in patients with clinical high risk for psychotic disorders. Span J Psychiatry Ment Health. 18(2):133–140. https://doi.org/10.1016/j.sjpmh.2024.09.001

Stafford MR, Jackson H, Mayo-Wilson E, Morrison AP, Kendall T. 2013. Early interventions to prevent psychosis: systematic review and meta-analysis. BMJ. 346(jan18 1):f185-f185. https://doi.org/10.1136/bmj.f185

Stain HJ et al. 2016. A randomised controlled trial of cognitive behaviour therapy versus non-directive reflective listening for young people at ultra high risk of developing psychosis: the detection and evaluation of psychological therapy (DEPTH) trial. Schizophr Res. 176(2-3):212-219. https://doi.org/10.1016/j.schres.2016.08.008

Stefana A et al. 2025. The mini-CAARMS: development and validation of a short version of the comprehensive assessment of at risk mental states to facilitate preventive psychiatry. Schizophr Bull. https://doi.org/10.1093/schbul/sbaf146

Stefanelli R et al. 2025. The semi-structured interview for bipolar at-risk states (SIBARS): psychometric properties and validation. J Affect Disord. 387:119529. https://doi.org/10.1016/j.jad.2025.119529

Stentebjerg-Olesen M, Pagsberg AK, Fink-Jensen A, Correll CU, Jeppesen P. 2016. Clinical characteristics and predictors of outcome of schizophrenia-spectrum psychosis in

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY

43

children and adolescents: a systematic review. J Child Adolesc Psychopharmacol. 26(5):410-427. https://doi.org/10.1089/cap.2015.0097

Stubbs B et al. 2016. Physical multimorbidity and psychosis: comprehensive cross sectional analysis including 242,952 people across 48 low- and middle-income countries. BMC Med. 14(1):189. https://doi.org/10.1186/s12916-016-0734-z

Stubbs B et al. 2018. EPA guidance on physical activity as a treatment for severe mental illness: a meta-review of the evidence and Position Statement from the European Psychiatric Association (EPA), supported by the International Organization of Physical Therapists in Mental Health (IOPTMH). Eur Psychiatry. 54:124–144. https://doi.org/10.1016/j.eurpsy.2018.07.004

Studerus E, Beck K, Fusar-Poli P, Riecher-Rossler A. 2020. Development and validation of a dynamic risk prediction model to forecast psychosis onset in patients at clinical high risk. Schizophr Bull. 46(2):252–260. https://doi.org/10.1093/schbul/sbz059

Sullivan SA et al. 2022. Developing and internally validating a prognostic model (P Risk) to improve the prediction of psychosis in a primary care population using electronic health records: the MAPPED study. Schizophr Res. 246:241–249. https://doi.org/10.1016/j.schres.2022.06.031

Taipale H et al. 2022. Representation and outcomes of individuals with schizophrenia seen in everyday practice who are ineligible for randomized clinical trials. JAMA Psychiatry. 79(3):210–218. https://doi.org/10.1001/jamapsychiatry.2021.3990

Tang Y et al. 2023. Visuospatial learning selectively enhanced by personalized transcranial magnetic stimulation over parieto-hippocampal network among patients at clinical high-risk for psychosis. Schizophr Bull. 49(4):923–932. https://doi.org/10.1093/schbul/sbad015

Td K, Rcw M JJ, JR, Lb G CW, BF, MN, By G, Bh E. 2018. White matter organization and cognition in patients at ultra-high risk for psychosis: preliminary results from the FOCUS-trial. Early Interv Psychiatry. 12.

Tsai CL et al. 2021. Effects of the health-awareness-strengthening lifestyle program in a randomized trial of young adults with an at-risk mental state. Int J Environ Res Public Health. 18(4):1959. https://doi.org/10.3390/jerph18041959

Valmaggia LR et al. 2015. Duration of untreated psychosis and need for admission in patients who engage with mental health services in the prodromal phase. Br J Psychiatry. 207(2):130-134. https://doi.org/10.1192/bjp.bp.114.150623

Valmaggia LR et al. 2009. Economic impact of early intervention in people at high risk of psychosis. Psychol Med. 39(10):1617-1626. https://doi.org/10.1017/S0033291709005613

van der Gaag M et al. 2012. Cognitive behavioral therapy for subjects at ultrahigh risk for developing psychosis: a randomized controlled clinical trial. Schizophr Bull. 38(6):1180–1188. https://doi.org/10.1093/schbul/sbs105

van der Gaag M et al. 2013. Preventing a first episode of psychosis: meta-analysis of randomized controlled prevention trials of 12 month and longer-term follow-ups. Schizophr Res. 149(1-3):56–62. https://doi.org/10.1016/j.schres.2013.07.004

Vissink CE et al. 2022. Structural brain volumes of individuals at clinical high risk for psychosis: a meta-analysis. Biol Psychiatry Glob Open Sci. 2(2):147–152. https://doi.org/10.1016/j.bpsgos.2021.09.002

Waite F et al. 2023. A targeted psychological treatment for sleep problems in young people at ultra-high risk of psychosis in England (SleepWell): a parallel group, single-blind, randomised controlled feasibility trial. Lancet Psychiatry. 10(9):706-718. https://doi.org/10.1016/S2215-0366(23) 00203-1

Walter A et al. 2016. Hippocampal volume in subjects at clinical high-risk for psychosis: a systematic review and meta-analysis. Neurosci Biobehav Rev. 71:680–690. https://doi.org/10.1016/j.neubiorev.2016.10.007

Wannan CMJ et al. 2024. Accelerating medicines partnership(r) schizophrenia (AMP(R) SCZ): rationale and study design of the largest global prospective cohort study of clinical high risk for psychosis. Schizophr Bull. 50(3):496-512. https://doi.org/10.1093/schbul/sbae011

Wasserthal S et al. 2024. N-acetylcysteine and a specialized preventive intervention for individuals at high risk for psychosis: a randomized double-blind multicenter trial. Schizophr Bull Open. 5(1):sgae005. https://doi.org/10.1093/schizbullopen/sgae005

Wenneberg C et al. 2020. Cerebral glutamate and GABA levels in high-risk of psychosis states: a focused review and meta-analysis of (1)H-MRS studies. Schizophr Res. 215:38–48. https://doi.org/10.1016/j.schres.2019.10.050

Wessels H et al. 2019. Predictors of treatment response to psychological interventions in people at clinical high risk of first-episode psychosis. Early Interv Psychiatry. 13(1):120–127. https://doi.org/10.1111/eip.12460

Wijnen BFM et al. 2020. Designing and testing of a health-economic markov model for prevention and treatment of ear-ly psychosis. Expert Rev Pharmacoecon Outcomes Res. 20(3): 269-279. https://doi.org/10.1080/14737167.2019.1632194

Wilson R et al. 2019. Cannabidiol attenuates insular dysfunction during motivational salience processing in subjects at clinical high risk for psychosis. Transl Psychiatry. 9(1):203. https://doi.org/10.1038/s41398-019-0534-2

Wolff RF, P. et al. 2019. PROBAST: a tool to assess the risk of bias and applicability of prediction model studies. Ann Intern Med. 170(1):51-58. https://doi.org/10.7326/M18-1376

Woods S et al. 2017. 112. Effects of ziprasidone versus place-bo in patients at clinical high risk for psychosis. Schizophr Bull. p. S58.

Woods SW, 3rd et al. 2021. Counterpoint. Early intervention for psychosis risk syndromes: minimizing risk and maximizing benefit. Schizophr Res. 227:10–17. https://doi.org/10.1016/j.schres.2020.04.020

Woods SW et al. 2024. Development of the PSYCHS: positive SYmptoms and Diagnostic Criteria for the CAARMS harmonized with the SIPS. Early Interv Psychiatry. 18(4):255–272. https://doi.org/10.1111/eip.13457

Woods SW et al. 2013. Glycine treatment of the risk syndrome for psychosis: report of two pilot studies. Eur Neuropsychopharmacol. 23(8):931–940. https://doi.org/10.1016/j.euroneuro.2012.09.008

Worthington MA et al. 2022. Individualized prediction of pro-dromal symptom remission for youth at clinical high risk for psychosis. Schizophr Bull. 48(2):395–404. https://doi.org/10.1093/schbul/sbab115

Worthington MA et al. 2021. Incorporating cortisol into the NAPLS2 individualized risk calculator for prediction of psychosis. Schizophr Res. 227:95–100. https://doi.org/10.1016/j.schres.2020.09.022

44 M. SOLMI ET AL.

Yassin W et al. 2020. Machine-learning classification using neuroimaging data in schizophrenia, autism, ultra-high risk and first-episode psychosis. Transl Psychiatry. 10(1):278. https://doi.org/10.1038/s41398-020-00965-5

Yoviene Sykes LA et al. 2020. Predictive validity of conversion from the clinical high risk syndrome to frank psychosis. Schizophr Res. 216:184–191. https://doi.org/10.1016/j.schres.2019.12.002

Yung AR et al. 2011. Randomized controlled trial of interventions for young people at ultra high risk for psychosis: 6-month analysis. J Clin Psychiatry. 72(4):430–440. https://doi.org/10.4088/JCP.08m04979ora

Yung AR et al. 2005. Mapping the onset of psychosis: the Comprehensive assessment of at-risk mental states. Aust N. Z J Psychiatry. 39(11-12):964–971. https://doi.org/10.1080/j.1440-1614.2005.01714.x

Zarogianni E et al. 2019. Individualized prediction of psychosis in subjects with an at-risk mental state. Schizophr Res. 214:18–23. https://doi.org/10.1016/j.schres.2017.08.061

Zeng J et al. 2023. A meta-analysis of neural correlates of reward anticipation in individuals at clinical risk for schizophrenia. Int J Neuropsychopharmacol. 26(4):280–293. https://doi.org/10.1093/ijnp/pyad009

Zhang T et al. 2019. Prediction of psychosis in prodrome: development and validation of a simple, personalized risk

calculator. Psychol Med. 49(12):1990–1998. https://doi.org/10.1017/S0033291718002738

Zhao J et al. 2023. EMDR versus waiting list in individuals at clinical high risk for psychosis with post-traumatic stress symptoms: a randomized controlled trial. Schizophr Res. 256:1-7. https://doi.org/10.1016/j.schres.2023.04.003

Zhao Y et al. 2022. Cortical thickness abnormalities at different stages of the illness course in schizophrenia: a systematic review and meta-analysis. JAMA Psychiatry. 79(6):560-570. https://doi.org/10.1001/jamapsychiatry. 2022.0799

Zheng Y et al. 2022. Cognitive behavioral therapy for prodromal stage of psychosis-outcomes for transition, functioning, distress, and quality of life: a systematic review and meta-analysis. Schizophr Bull. 48(1):8–19. https://doi.org/10.1093/schbul/sbab044

Zhu H et al. 2023. Effect of Shi-Zhen-An-Shen herbal formula granule in the treatment of young people at ultra-high risk for psychosis: a pilot study. Front Psychiatry. 14:1160452. https://doi.org/10.3389/fpsyt.2023.1160452

Zhu Y et al. 2024. Using brain structural neuroimaging measures to predict psychosis onset for individuals at clinical high-risk. Mol Psychiatry. 29(5):1465–1477. https://doi.org/10.1038/s41380-024-02426-7

相关指南

• 2024 ACG指南：急性胰腺炎的管理 · 消化内科 2024
• 2025 ESPEN指南：外科手术中的临床营养（更新版） · 护理与营养 2025
• 标准与规范丨中国重症医学科建设和发展指南（2025版） · 重症医学科
• 2025 IDSA指南：复杂性尿路感染的管理和治疗—复杂性尿路感染的抗生素治疗选择 · 感染与传染病 2025
• 重症凝血病标准化评估中国专家共识(2025版) · 血液内科 2025
• 2024 ACVC／EHRA／HFA临床共 识声明：急性心力衰竭中的室性心律失常 · 心血管内科 2024
• （2024）成人腹腔高压和腹腔间隔室综合征诊治急诊专家共识 · 麻醉与急重症 2024
• 免疫抑制宿主重症肺炎诊断与治疗专家共识 · 呼吸内科 2026