Giada Tripoli

30 Years on: How the Neurodevelopmental Hypothesis of Schizophrenia Morphed Into the Developmental Risk Factor Model of Psychosis

At its re-birth 30 years ago, the neurodevelopment hypothesis of schizophrenia focussed on aberrant genes and early neural hazards, but then it grew to include ideas concerning aberrant synaptic pruning in adolescence. The hypothesis had its own stormy development and it endured some difficult teenage years when a resurgence of interest in neurodegeneration threatened its survival. In early adult life, it over-reached itself with some reductionists claiming that schizophrenia was simply a neurodevelopmental disease. However, by age 30, the hypothesis has matured sufficiently to incorporated childhood and adult adversity, urban living and migration, as well as heavy cannabis use, as important risk factors. Thus, it morphed into the developmental risk factor model of psychosis and integrated new evidence concerning dysregulated striatal dopamine as the final step on the pathway linking risk factors to psychotic symptoms.

35.4 A PUBLIC HEALTH APPROACH TO THE PREVENTION OF PSYCHOSIS

Abstract Background The main attempt to prevent the development of psychosis has been through clinics for people at clinical high risk. Such an approach is useful for research but can never reach the majority of individuals who will become psychotic. Biological markers could be used to identify individuals with unusual vulnerabilities e.g. those with copy number variations such as VCFS. However, identifying the with such markers is unlikely to impact on the majority of cases, and as yet no useful interventions are available. How therefore to prevent psychosis? Methods Data will be presented from 3 studies of first onset psychosis (FEP) which used similar methods of ascertainment and assessment of cases and controls; AESOP and GAP from South London and the EU-GEI across 16 sites in 5 European countries. Results The identified risk factors for psychosis were the polygenic risk score for schizophrenia, childhood abuse, living in a city, being from an ethnic minority, drug abuse, adverse life events. Clearly, reducing some of these (e.g. urbanicity or migration) is not within the powers of psychiatrists. The GAP study showed that the polygenic risk score accounted for the greatest variance in caseness; those with scores in the highest quintile were 7 times more likely to be a psychotic case than those in those lowest quintile. The GAP study also gave estimates of the population attributable fraction (PAF): these indicated that if no one was exposed to child abuse and use of high potency cannabis, then 16% and 24% respectively of psychosis in South London could be prevented. The EU-GEI study showed striking differences in the incidence of psychosis between Northern and Southern Europe; data will be prevented concerning the contribution of risk factors, especially cannabis use, to this. Discussion The knowledge that schizophrenia is the extreme of a continuum of psychosis has important implications for prevention. Preventive approaches to hypertension or obesity do not focus on identifying individuals carrying biological markers; rather they encourage members of the general population to take exercise and reduce their calorie intake. A similar approach should be adopted for psychosis. In the long-term attempts to reduce risk factors should be made e.g. addressing psychotogenic aspects of city living or by decreasing discrimination of ethnic minorities. This will be difficult. However, an obvious place to start is by attempting to influence society’s patterns of consumption of high-potency cannabis. Unfortunately, public policy in the US and certain other countries appears to be moving in the opposite direction with increases in consumption and potency. Are these countries sleep-walking to more psychosis?

T110. FIRST EPISODE PSYCHOTIC PATIENTS WITH A HISTORY OF FREQUENT CANNABIS USE EXPRESS MORE POSITIVE SYMPTOMS AT ILLNESS ONSET THAN THOSE WHO NEVER USED CANNABIS

Abstract Background Robust evidence has demonstrated that cannabis use increases the risk to develop psychotic disorders. However, a limited number of studies have investigated if and how cannabis use influences psychopathology profiles at first episode psychosis (FEP). Based on the evidence that dopamine dysfunction contributes to explain positive symptoms in psychosis, and that the main cannabis’ psychoactive component, Δ9-Tetrahydrocannabinol (THC), modulates the dopamine system, we hypothesise that: 1) positive symptoms at FEP are more common among psychotic patients who used cannabis compared with never users; 2) this association is a dose-response relationship. Methods We analyzed a sample of 1130 FEP patients as part of the EUGEI study, recruited across six countries. The MRC Socio-demographic Schedule was used to collect sociodemographic information. Psychopathology was assessed with the OPerational CRITeria (OPCRIT), and symptom items were analyzed using Mplus to estimate a multidimensional model of psychosis. The Cannabis Experience Questionnaire modified version (CEQmv) was administered to collect information on cannabis, and different patterns of use were computed based on frequency of consumption and type of cannabis, as a proxy of exposure to THC. Results The lifetime rate of cannabis use was 63%. Fifty-five percent of cannabis users consumed mostly high-potency cannabis, and 46% showed a daily frequency. Mixed-effects linear regression revealed that frequency of cannabis use was associated with the positive symptom dimension score. Daily users of high-potency cannabis presented with the strongest association (Β=0.19, 95%CI=0.02–0.38), even after gender, age, ethnicity, other drug use, and study site were controlled for. Discussion Our results show that patients with a history of daily use of high potency cannabis express more positive symptoms at psychosis onset, even after taking into account other substance use and relevant sociodemographic factors.

F99. FIRST EPISODE PSYCHOSIS PATIENTS WHO USED CANNABIS DEVELOP THEIR ILLNESS AT A SIGNIFICANTLY YOUNGER AGE THAN THOSE WHO NEVER USED CONSISTENTLY ACROSS EUROPE AND BRAZIL

Abstract Background Patients presenting to psychiatric services with their first episode of psychosis (FEP) report higher rates of previous cannabis use than the general population (Donoghue et al., 2011; Myles, Myles and Large, 2016). Evidence suggested that patients suffering from psychosis with a history of cannabis use have an earlier age of onset of psychosis (AOP) than those who never used it (Di Forti et al., 2013). We aim to investigate if the reported association between use of cannabis and AOP is consistent across different countries, once having taken into account different patterns of cannabis use (i.e. frequency of use and age at first use). Methods We analysed data on patterns of lifetime cannabis use and AOP from FEP=1,149 (61.7% males) from 5 European countries and Brazil part of the European network of national schizophrenia networks studying European Gene-Environment-Interaction (EUGEI) study. Patients met ICD-10 criteria for psychosis, ascertained by using OPCRIT (McGuffin et al., 1991). The CEQmv (Di Forti et al., 2009) further modified for the EUGEI study, was used to collect data on lifetime frequency of cannabis use (never used/used at least once but less than daily/ everyday use) and age at first use in years (then dichotomized according to mean age at first use ≤15 years or ≥16 years). We used two ANOVAs: age of onset was used as the outcome variable and frequency of cannabis use and age of first use were respectively entered as independent predictors, along with country, gender and self-ascribed ethnicity. Results 63.3% of our sample used cannabis at least once in lifetime. Among those who used cannabis in their lifetime, mean age at first use was 16.8 years (sd=4.6) and median age was 16 years, 42.3% tried first time cannabis at 15 years or before, 57.7% at 16 years or older. Patients who smoked cannabis on a recreational basis (mean age 29.0; contrast=5.8, CI 95% 4.3, 7.2, p<0.001) and on a daily basis (mean age 26.6; contrast=2.4, CI 95% 0.9, 3.9, p=0.001) had lower age of onset than not users patients (mean age 34.8) across all countries, once have taken into account gender and ethnicity Only, those who started using cannabis ≤15 years had an earlier age of onset (25.5 years) than those who started at their 16 years or later (29.5 years), (F(1,683)=37.3, p<0.001). This relationship was the same across different countries (p=0.968), and independently influenced by ethnicity (F(5, 683)=2.3, p=0.03) but not by gender (p=0.057). Discussion Our results suggest a generalizable across country and specific effect of frequency of use and early age at first cannabis use on significantly anticipate age of psychosis onset in First episode Psychosis patients.

S77. JUMPING TO CONCLUSIONS AND FACIAL EMOTION RECOGNITION IMPAIRMENT IN FIRST EPISODE PSYCHOSIS ACROSS EUROPE

Abstract Background Jumping to conclusions (JTC) is a well-established reasoning and data gathering bias found in patients with psychosis even at illness onset (First Episode Psychosis, FEP). Preliminary work in this field focused primarily on the association with delusions, although jumping to conclusions has also been found in non-deluded schizophrenia patients after remission, and in individual with at risk mental state. Moreover, psychotic patients tend to show impairments in social cognition, struggling in identifying, processing and interpreting social clues. Deficits in facial emotion recognition (FER) – a key component of the construct – represent a well-replicated finding in schizophrenia. Furthermore, deficits in global facial affect recognition have been found in FEP with the same severity as at further stages, especially for anger recognition. The present study aims to measure JTC and FER bias in a sample of FEP recruited across 5 European countries, compared with healthy controls. Methods Data on JTC (Beads task 60:40), FER (Degraded Facial Recognition task – DFAR) and socio-demographics have been analysed in a sample of 643 FEP and 1019 population controls recruited as part as the EU-GEI study across UK, Netherlands, France, Spain, and Italy. IQ scores were used to exclude cases and controls with current IQ<70 (N=171) from JTC analysis and a score <41 (N=384) on the Benton Facial Recognition test for the analysis on DFAR. Logistic regression model was applied to predict case/control status using 1) JTC and 2) DFAR as predictive variables controlling for age, gender and country. Results We showed that the presence of JTC bias varies across different countries both in cases (χ2=23.77 p<0.001) and controls groups (χ2=14.01 p=0.007). Logistic regression analyses revealed JTC to be a significant predictor of case/control status (Adj OR=1.88 CI 95%=1.43–2.29 p<0.001). As well as JTC, FER differed over Europe in both groups (FEP, total: F=17.37, p<0.001; neutral: F=12.4, p<0.001; happy: F=25.62, p<0.001; frightened: F=8.78, p<0.001; angry: F=5.48, p<0.001. Controls, total: F=23.06, p<0.001; neutral: F=21.72, p<0.001; happy: F=21.74, p<0.001; frightened: F=14.14, p<0.001; angry: F=12.49, p<0.001). Logistic regression analyses revealed all DFAR scores, except for happy emotions, to be negatively associated with case/control status (total: B=-.0182 p=0.001; neutral: B=-.054 p=0.003; happy: B=-.0196 p=0.2; frightened: B= -.065 p<0.001; angry: B=-.030 p=0.04). Discussion This study supports the evidence that 1) FEP patients are more likely to present JTC and FER impairments than controls; 2) cognition and social cognition might represent transcultural features of psychotic disorders.

5.4 BIOLOGICAL AND EPIDEMIOLOGICAL EXAMINATION OF TRANSDIAGNOSTIC AND SPECIFIC SYMPTOM DIMENSIONS AT PSYCHOSIS ONSET: FINDINGS FROM THE EUGEI STUDY

Abstract Background Current diagnostic models of psychosis have been questioned since Kraepelin’s original dichotomy of dementia praecox and manic depression. Indeed, increasing evidence has suggested that a dimensional approach might be a valid alternative platform for research. However, while an increasing number of studies have investigated how environmental risk factors for affective and non-affective psychosis map onto symptom dimensions, only a few have examined these dimensions in relation to genetic variants as summarised by Polygenic Risk Score (PRS). Furthermore, no studies have examined the putative effect of PRS for Schizophrenia (SZ), Bipolar Disorder (BP), and Major Depressive Disorder (MDD) on previously identified general and specific symptom dimensions. At the same time, only one study has investigated how symptoms vary according to epidemiological factors such as living in urban neighbourhoods. The objectives of this study were to: 1) test whether a bi-factor model statistically fits the conceptualization of psychosis as composed of general and specific dimensions; 2) examine the extent to which SZ, BP, and MDD PRSs explain the phenotypic variance due to general and specific dimensions; 3) test the hypothesis that the general psychosis dimension would be more severe in highly urban environments. Methods We used clinical and epidemiological data from the EUropean network of national schizophrenia networks studying Gene-Environment Interactions (EUGEI) study, including 2322 First Episode Psychosis (FEP) patients recruited in 17 sites across 6 countries. Genetic variants were collectively analyzed for 800 individuals. The following analysis steps were performed: 1) Psychopathology items were analysed using multidimensional item response modelling in MPlus to estimate unidimensional, multidimensional, and bi-factor models of psychosis. Model fit statistics included Log-Likelihood, and Akaike and Bayesian Information Criteria to compare these models. 2) SZ, BP, and MDD PRSs for general and specific dimensions were built using PRSice. Summary statistics from large case-control mega-analyses from the Psychiatric Genomics Consortium were used as base data sets and general and specific dimension scores were used as discovery data sets. Individuals’ number of risk alleles in the discovery sample was weighted by the log odds ratio from the base samples, accounting for population stratification, and summed into the three PRSs. 3) Multilevel regression analysis was used in STATA 14 to examine the variance in general dimension due to the population density levels across the sites. Results A bi-factor solution, composed of one general and five specific symptom dimensions, showed the best model fit statistics. Higher SZ PRS score was associated with higher scores on positive dimensions (β= 0.27, t=2.11, p<0.05); higher BP PRS was associated with higher scores on mania dimension (β= 0.17, t=2.11, p<0.05); higher MDD PRS was associated with lower scores on negative dimension (β= -0.31, t=-2.25, p<0.05). No trends of association were found for SZ, BP, or MDD PRSs and the general psychosis dimension. The transdiagnostic symptom dimension score was elevated in people living in more densely populated sites (η2=0.077, 95% CI 0.057–0.098). Discussion Our results suggest that a) symptom dimension structure at FEP is best represented by the bi-factor model; b) in FEP patients, there is a trend of associations between SZ PRS and positive dimension, and between BP PRS and mania dimension; and c) elevated level of transdiagnostic symptomatology was observed in more densely populated sites.

F68. PREMORBID IQ, EDUCATIONAL LEVEL AND JUMPING TO CONCLUSIONS AS PREDICTORS OF CLINICAL OUTCOME AT FIRST ONSET OF PSYCHOSIS OVER THE NEXT 4 YEARS: THE GAP STUDY

Abstract Background Cognition and more recently social cognition, have been shown to be a strong predictor of clinical and functional outcome in psychosis. Jumping to Conclusions (JTC), which is defined as the proneness to require less information before forming beliefs or making a decision, has been related to the formation and maintenance of delusions. However, its relevance to longer-term outcome is unclear. On the other hand, there is evidence in the literature to suggest differences of patterns in clinical outcome and service based ethnicity. Using data from the GAP case-control study of first-episode psychosis (FEP), we set out to test whether the premorbid IQ, educational level and presence of JTC would predict poor clinical outcome at 4 year controlling for ethnicity. Methods 431 FEP patients were assessed with the positive and negative syndrome scale (PANSS) and Global Assessment of Functioning (GAF). Premorbid IQ was measured by the National Adult Reading Test (NART) scale, probabilistic reasoning “Beads” task was applied and educational levels were recorded alongside with socio-occupational variables at the time of recruitment. Follow-up data over an average period of 4 years were obtained from the electronic psychiatric clinical records in the South London and Maudsley NHS Foundation Trust (SLaM); including items concerning clinical course and outcomes (remission, intervention of police, use of involuntary treatment – the Mental Health Act (MHA) -, and inpatient days). We build different regression models using separately premorbid IQ, education level and JTC as predictors for each clinical outcome, both unadjusted and adjusted by ethnicity, age and gender. Results Higher educational level was predictor of clinical remission [adjusted OR=1.9, 95% confidence interval (CI) 1.2–3, p=0.005]. FEP who presented JTC at baseline were more likely during the follow up period to be detained under the MHA [adjusted OR=11.23, 95% confidence interval (CI) 2.64–47.76, p=0.001], require intervention by the police (adjusted OR=10.76, 95% CI 2.4–48.26, p=0.002) and have longer admissions (adjusted IRR=4.04, 95% CI 1.43–11.36, p=0.008). We couldn’t find any predictor effect for clinical outcome for premorbid IQ. The association with level of education and JTC was not accounted for by socio-demographic variables including ethnicity. Discussion Although we did not find association with premorbid IQ, educational level as indirect proxy of neurocognition showed a predictor effect for clinical remission. JTC in FEP is associated with serious subsequent consequences in terms of social disturbance and a poor therapeutic alliance. Our findings raise the question of whether the implementation of specific interventions to reduce JTC, such as Metacognition Training, may be a useful addition in early psychosis intervention programs.

O2.1. FIRST EPISODE PSYCHOSIS PATIENTS ACROSS EUROPE DIFFER IN INTELLECTUAL QUOTIENT (IQ) AND EXPOSURE TO ENVIRONMENTAL HAZARDS

Abstract Background Children who later develop Schizophrenia on average are more likely to present with lower IQ; this has been considered evidence for the neurodevelopmental theory of schizophrenia. Though, recent studies have shown that first episode psychosis patients with a history of cannabis use have significantly higher premorbid and current IQ compared to those who never used it. This suggests that abnormal early neurodevelopment does not explain the aetiology of all cases of Schizophrenia, leaving space to environmental hazards. The present study aims to: investigate differences in IQ, as a marker of neurodevelopment, and in exposure to environmental risk factors in a large sample first episode psychosis patients recruited across five different European countries, in comparison with their respective control groups. Methods We analysed data on IQ, socio-demographics and cannabis use from FEP=705 (51.1 % males) and healthy controls=1.034 (48.9 % males), as part of the European network of national schizophrenia networks studying European Gene-Environment-Interaction (EUGEI) study. Patients met ICD-10 criteria for psychosis, ascertained by using OPCRIT (McGuffin et al., 1991). The CEQmv(Di Forti et al., 2009) further modified for the EUGEI study, was used to collect data on cannabis use. We used ANOVAs where IQ was used as the outcome variable and case/control status and Country were respectively entered as independent predictors, along with other predictors. Results Case-control status (F (1,1.484)=133.1, p<0.001) and Country (F (4, 1.484)=32.1, p<0.001) resulted in interaction in predicting IQ after controlling for gender, age, ethnicity, education, occupation relationship and living status. That means, being a case and being from France (mean IQ=73.4), Spain (mean IQ=74.6) and Italy (mean IQ=75.4) was associated with the lowest IQ (F (4,1.484)=3.7, p=0.004), compared with cases from UK (mean IQ=87.1) and Holland (mean IQ=85.5). Among controls the pattern was similar but not significant. We then grouped countries as North- (UK and Holland) and South – Europe (Italy, Spain, France) and we compared the presence of the main risk factors between the two groups. Both patients and controls from the northern part of Europe, were more likely to be from other ethnicities (chi2(2)=93.3, p<0.001) and living alone (chi2(2)=39.6, p<0.001) than patients and controls from the southern part, who were, for instance, more likely to be married (chi2(2)=34.1, p=0.007). There were no differences in education, nor in gender distribution between cases from the north and from the south of Europe and cases from the south, but not controls, were more likely to be employed than patients from the North (chi2(1)=19.1, p<0.001). A significant difference emerged in patterns of cannabis use, that resulted more dangerous in the northern part, where cases, but not controls, were more likely to have used cannabis (chi2 (1)=18.2, p<0.001) on a daily basis (chi2(2)=17.6, p<0.001) with a higher concentration in THC (chi2 (2)=43.8, p<0.001) and before their 15 years (chi2 (2)=20.8, p<0.001), compared to patients from the south. Discussion Our findings on the higher IQ reported in the first episode psychosis patients from northern Europe sites might indicate this as a group with less neurodevelopmental abnormalities and more likely to have developed Psychosis because of adverse social environment and more harmful pattern of cannabis use, compared to patients from the southern countries.

O12.4. SOME OF THE INDIVIDUAL DIFFERENCES IN RISK TO DEVELOP PSYCHOSIS AMONG CANNABIS USERS CAN BE EXPLAINED BY WHERE THEY LIVE AND BY THEIR AGE AT FIRST USE

Abstract Background Cannabis use remains the most widely used recreational drug worldwide. Following from several USA states legalisation policies, European countries are reconsidering their cannabis laws. While a significant amount of Epidemiological evidence has reported that cannabis use increases the risk of psychosis it is still unclear: 1) what underpins individual differences in developing a psychotic disorder following cannabis use; 2) if variations in availability of cannabis have affected rate of Psychotic disorders across Europe. Methods Using detailed data on lifetime pattern of cannabis use from the EUGEI first episode case-control study (N=2300) and the available Incidence rates of Psychosis calculated for each European site of the same study, we aim 1) to estimate if differences in age at first use, especially of high potency cannabis among cannabis users resulted in differences in their probability to develop psychosis across the study sites; 2) to calculate the proportion of new cases of psychosis attributable to early adolescence-high Potency cannabis in the 5 countries; 3) to relate data on prevalence of cannabis use in each study site with the corresponding Incidence rates for psychotic disorders. Results Cannabis users starting using cannabis at age 15 and younger who live in those EU countries where high potency cannabis is available have the highest probability to develop psychosis, compared to never users (Adj ORs from 2.6–5.9; p<0.01). Moreover, the proportion of new cases of Psychosis attributable to heavy use started in adolescence was between 20% and 37%. Finally, the correlation between lifetime use of cannabis in population controls from the study sites was significantly correlated with the corresponding incidence rates for Psychosis (r=0.6; p<0.001) Discussion Before Europe rushes into the USA legalisation “moda” more public education effort might need to be invested in reducing the use of high potency type of cannabis among young adolescents. The latter could lead to a significant reduction in the proportion of new cases of psychosis across Europe.

Cumulative social disadvantage and psychosis: findings from a southern Italy case-control study

IEPA 10th International Conference on Early Intervention in Mental Health – “Looking Back, Moving Forward” Milan, Italy, 20th – 22nd October 2016