There is limited research in adolescent at risk for psychosis. The new criteria of Attenuated Psychosis Syndrome (APS) of Diagnostic and Statistical Manual of Mental Disorders- 5 (DSM-5) have not been validated.

The aims of this study were to: 1) characterize adolescent’s profile with APS (DSM-5 APS) compared to adolescents with early onset psychosis (EOP) and with other psychiatric disorders (non-APS); 2) to estimate their long-term risk of transition to psychosis and prognostic accuracy of DSM-5 APS.

243 adolescents, aged 12-17, were included (October 2012- July 2019) and dived in three sub-groups (110 DSM-5 APS, 31 EOP, 102 non-APS). All underwent a comprehensive assessment evaluating: sociodemographic characteristics, family and personal history of any DSM-5 psychiatric disorders, psychopathological assessment and level of functioning. An annual follow-up evaluation was carried out (up to 7 years) including a clinical interview to investigate DSM-5 criteria for transition to psychosis.

DSM-5 APS adolescents had on average higher comorbid disorders (2.3) and intermediate psychopathological and functioning profile between non-APS/EOP. The cumulative risk of transition at 1,2,3, 4-5 years was 13%, 17%, 24.2%, 26.8% and 26.8% in DSM-5 APS group, 0%, 0%, 3.2%, 3.2% and 3.2% in the non-APS. The 5-year prognostic accuracy of the DSM-5 APS in adolescent was adequate (Area Under the Curve=0.77) with high sensitivity (91.3%) and suboptimal specificity (63.2%).

The DSM-5 APS diagnosis can be used to detect help-seeking adolescents at risk of psychosis and predict their long-term outcomes, leading the way to new preventive approaches.

The authors declare that they do not have a significant financial interest, consultancy or other relationship with products, manufacturer(s) of products or providers of services related to this abstrac.

Prevalence and covariates of subclinical psychosis have gained increased interest in the context of early identification and treatment of persons at risk for psychosis.

We analysed 9829 adults representative of the general population within the canton of Zurich, Switzerland. Two psychosis syndromes, derived from the SCL-90-R, were applied: ‘schizotypal signs’ and ‘schizophrenia nuclear symptoms’.

Only a few subjects (13.2%) reported no schizotypal signs. While 33.2% of subjects indicated mild signs, only a small proportion (3.7%) reported severe signs. A very common outcome was no ‘schizophrenia nuclear symptoms’ (70.6%). Although 13.5% of the participants reported mild symptoms, severe nuclear symptoms were very rare (0.5%). Because these two syndromes were only moderately correlated (r = 0.43), we were able to establish sufficiently distinct symptom clusters. Schizotypal signs were more closely connected to distress than was schizophrenia nuclear symptoms, even though their distribution types were similar. Both syndromes were associated with several covariates, such as alcohol and tobacco use, being unmarried, low education level, psychopathological distress and low subjective well-being.

Subclinical psychosis symptoms are quite frequent in the general population but, for the most part, are not very pronounced. In particular, our data support the notion of a continuous Wald distribution of psychotic symptoms in the general population. Our findings have enabled us to confirm the usefulness of these syndromes as previously assessed in other independent community samples. Both can appropriately be associated with well-known risk factors of schizophrenia.