Psychosis Risk

Higher cannabis use is associated with increased psychosis risk (OR 3,90 for heavy users vs. non-users); dose-response relationship confirmed.

Meta-analysis across 10 studies (n≈66.816) on the dose-response relationship between cannabis use and psychosis risk; OR=3.90 (95% CI 2.84–5.34) for daily vs. non-use, OR=1.97 (1.68–2.31) for any use vs. never use; clear evidence for increased psychosis risk with heavier cannabis use.

50 to 70 percent achieved symptomatic remission within six months, but functional recovery was inconsistent; continued cannabis use was the strongest predictor of unfavorable courses.

Systematic review on cannabis-induced psychosis across k=33 studies (n=18.117, mean follow-up 24 months). While 50–70% achieved symptomatic remission within 6 months, functional recovery showed substantial heterogeneity with frequent symptom-function dissociation. Continued cannabis use was the strongest predictor of poor outcomes across all domains (persistence rate 35–65%); abstinence was associated with significant benefits (functional recovery effect size 1.26). Only 15% of studies used person-centered trajectory modeling, which identified five distinct recovery profiles.

Continued cannabis use after psychosis onset leads to higher relapse rates, longer hospital stays and more severe positive symptoms than abstinence or non-use.

Systematic review and meta-analysis across k=24 studies (n=16.565) on continued vs. discontinued cannabis use after psychosis onset (follow-up ≥6 months). Continued users (CC) showed increased relapse risk vs. non-users (NC) (d=0.36, 95% CI 0.22–0.50, p<0.0001) and longer inpatient stays (d=0.36, 95% CI 0.13–0.58, p=0.02); also increased vs. abstinent users (DC) (d=0.28, 95% CI 0.12–0.44, p=0.0005). Cannabis abstinence was not associated with relapse (d=0.02, 95% CI -0.12–0.15, p=0.82). Meta-regression: CC vs. DC significant (p=0.04).

Meta-analysis on the prevalence of schizophrenia-spectrum (20%) and bipolar disorders (5%) after cannabis-induced psychosis; no intervention assessment.

Systematic review + meta-analysis across k=13 studies (n=7.515) on cannabis-induced psychosis and subsequent diagnoses. 20% (95% CI: 15,8–29,5%) later developed schizophrenia-spectrum disorder, 5% (95% CI: 2,7–6,9%) bipolar disorder, 63% (95% CI: 26,8–90,5%) unspecified diagnosis. Patients had a 76% lower risk for bipolar disorder vs. schizophrenia-spectrum; unspecified diagnosis showed ARR=2,52 (95% CI: 1,03–6,15) vs. schizophrenia-spectrum alone.

Meta-analysis shows significantly increased psychosis risk with weekly or more frequent cannabis use (RR=1,35-1,76), but no significant risk with less frequent use.

Systematic review + dose-response meta-analysis across k=10 studies (3 cohorts, 7 case-control, n=7.390, age 12–65 years); significant log-linear dose-response relationship between cannabis use frequency and psychosis risk. Restricted cubic splines model shows risk thresholds: RR=1,35 (95% CI 1,19–1,52) with weekly use, RR=1,76 (95% CI 1,47–2,12) with daily use; no significant risk with <weekly use (RR=1,01 yearly, RR=1,10 monthly).

Cannabis use increases the risk of non-adherence to antipsychotics (OR 2,46); cannabis abstinence is associated with better adherence.

Systematic review and meta-analysis across k=15 observational studies (n=3.678) on cannabis use and medication adherence in psychosis patients. Cannabis users showed increased risk of non-adherence vs. non-users (OR=2.46). At follow-up: current users vs. non-users OR=5.79 (n=175), current users vs. former users OR=5.5 (n=192), no difference between former users and non-users (OR=1.12, n=187). Cannabis abstinence normalises adherence.

No significant association between cannabis use and duration of untreated psychosis, only a non-significant trend towards shorter DUP.

SR + meta-analysis across k=9 studies (n=1.726 first-episode psychosis patients, of whom 39% cannabis users); cannabis use showed a trend towards shorter duration of untreated psychosis (DUP), however the meta-analysis found no statistically significant association between DUP and cannabis use.

Lifetime cannabis use was not significantly associated with psychosis transition (OR 1.14, p=0.37), but current cannabis abuse/dependence showed significant association (OR 1.75, p=0.01).

Systematic review + meta-analysis on cannabis in Ultra-High-Risk (UHR) for psychosis, k=7 prospective studies (n=1.171). Lifetime cannabis use NOT significantly associated with psychosis transition (OR=1,14, 95% CI 0,856–1,524, p=0,37). Current cannabis abuse/dependence (DSM-IV), however, significantly predictive (OR=1,75, 95% CI 1,135–2,710, p=0,01) — indication of a dose-response relationship.

Intravenous THC consistently induces psychotomimetic effects, including positive and negative symptoms; the effects are dose-dependent and reduced in frequent cannabis users.

IPD mega-analysis across k=10 double-blind, randomized, placebo-controlled crossover studies with n=400 i.v. THC infusions in healthy volunteers. Clinically meaningful increase in positive symptoms in 44,75% of infusions; positive symptoms positively associated with THC dose (β=11,13, p<0,001) and negatively associated with frequent cannabis use (β=-0,575, p<0,001). Strong correlation between positive symptoms and perceptual alterations (rs=0,514, p<0,001).

THC induces psychiatric symptoms with large effect sizes (positive, negative and general symptoms); CBD shows no consistent moderating effects.

Systematic review and meta-analysis across k=15 studies (acute THC administration) + 4 studies (CBD+THC) on psychiatric symptoms in healthy subjects. THC vs. placebo: significant increase in total symptom severity with large effect size (PANSS/BPRS total, positive and negative symptoms). Meta-analysis on 9 studies with 10 independent datasets. CBD moderation systematically reviewed (insufficient data for meta-analysis). PROSPERO CRD42019136674.

Cannabis use increases the risk of psychotic disorders in a dose-dependent manner, while evidence for affective outcomes is weaker and less consistent.

Systematic review (k=35 longitudinal population studies); pooled adjusted OR for psychotic outcome with ever cannabis use = 1,41 (95% CI 1,20–1,65); dose-response effect: OR=2,09 (95% CI 1,54–2,84) with frequent use. Evidence consistent with an independent causal effect of cannabis on psychosis risk; depression/anxiety outcomes less consistently supported.

Cannabis use increases the risk of schizophrenia approximately twofold, but is neither a sufficient nor a necessary condition for the development of psychosis.

Systematic evaluation of k=5 prospective population cohorts; cannabis use doubles the relative risk for schizophrenia (RR≈2,0); elimination of cannabis use would reduce schizophrenia incidence by ~8% (assuming causality). Cannabis = component cause, neither sufficient nor necessary. Stronger risk increase with adolescent first use (dose-dependent).

Cannabis use is associated with a significantly earlier onset of psychosis by an average of 2,70 years.

Meta-analysis of k=83 studies: cannabis use associated with an average of 2,70 years earlier onset of psychotic disorders vs. non-users (SMD=-0,414); for broadly defined substance use 2,00 years earlier (SMD=-0,315). Alcohol use showed no significant association. Proportion of cannabis users significantly explained the heterogeneity of effect sizes.

Cannabis use is associated with increased risk of psychosis onset, dose-dependent risk, earlier manifestation and, in existing psychosis, with more relapses, hospitalisations and stronger positive symptoms.

Umbrella review of k=26 systematic reviews and meta-analyses on cannabis and psychosis (2005-2016). Psychotic disorders occur more frequently in cannabis users; dose-dependent risk for psychotic disorders; earlier disease onset in cannabis users; increased relapse rates, more hospitalisations and more pronounced positive symptoms in psychotic patients with cannabis use.

Meta-analysis shows: cannabis initiation on average 6,3 years before psychosis onset; prevalence at First Episode Psychosis 33,7%; after treatment reduced risk for continued cannabis use (OR 0,56).

Meta-analysis across k=37 observational studies on cannabis use at first manifestation of psychosis. Interval between onset of regular cannabis use and psychosis onset: 6,3 years (SMD=1,56; 95% CI: 1,40–1,72). Prevalence of current cannabis use at first manifestation: 33,7% (95% CI: 31–39%). Odds for continued use 6 months to 10 years after first manifestation: OR=0,56 (95% CI: 0,40–0,79).

Regular cannabis use was associated with measurable neuroanatomical changes in individual brain regions, though the data situation was heterogeneous.

SR + meta-analysis (k=30 studies, 17 meta-analyses) on neuroanatomical changes in regular cannabis use vs. controls. Significantly smaller hippocampal volumes (SMD=0.14, 95% CI [0.02, 0.27], p=0.02, I²=74%) and medial orbitofrontal cortex (SMD=0.30, 95% CI [0.15, 0.45], p=0.0001, I²=51%) in users. Findings consistent with neuronal vulnerability of psychosis-associated circuits (reward, learning, memory).

Cannabis use in adolescence increases psychosis risk later in life (RR = 1,71), with several moderating factors such as age at cannabis onset, frequency of use, childhood trauma and genetic factors.

Systematic review + meta-analysis on adolescent cannabis use and psychosis risk; k=63 studies narrative, k=18 in meta-analysis. Adolescent cannabis use increased psychosis risk (RR=1,71; 95% CI 1,47-2,00; p<0,00001) and predicted earlier psychosis onset. Moderating factors: age of onset, frequency of use, childhood trauma, polysubstance use, genetic factors.

Cannabis use is robustly associated with a 32 months earlier onset of psychosis, independent of tobacco use and other confounders examined.

Meta-analysis on the association between cannabis use and earlier first onset of psychosis; cannabis users developed psychosis on average 32 months earlier than non-users (SMD=-0,399, 95% CI -0,493 to -0,306, p<0,001); finding robust against confounders (tobacco use, sex, diagnostic criteria); no corresponding effect for tobacco alone (SMD=0,002, p=0,974).

Cannabis use is consistently associated with increased schizotypy scores, with small to medium effect sizes.

Meta-analysis of k=29 cross-sectional studies; cannabis use vs. never use: increased schizotypy scores for all 4 dimensions (total, positive, negative, disorganised), effect sizes in the small to medium range, statistically significant (except negative score for current vs. other); cannabis use consistently associated with increased subclinical psychosis traits.

Cannabis use was not significantly associated with transition to psychosis (RR = 1,11, 95% CI = 0,89-1,37); limited evidence for other outcomes.

Systematic review on cannabis in clinical high risk (CHR) for psychosis, k=36 studies. Lifetime cannabis prevalence 48,7%, current use 25,8%, cannabis use disorder 14,9%. Pooled relative risk for psychosis transition RR=1,11 (95% CI 0,89–1,37, not significant). All cannabis prevalences with high heterogeneity (I²=75,7–92,8%).

Cannabis use is associated with negative psychotic outcomes in psychosis patients, also when controlling for important confounders.

Systematic review of long-term studies on cannabis use in psychosis patients; k=33 studies included. Overarching pattern of negative psychotic outcomes with cannabis use, also after controlling for important confounding variables. Psychosis patients should be informed about cannabis effects based on evidence; systematic evaluation and interventions to reduce use are recommended.

Co-exposure to childhood trauma and cannabis use shows synergistic effects on psychosis risk with odds ratios up to 20,9 and 31,0 respectively; combined exposure associated with earlier manifestation (2,9-3,6 years), more severe symptoms, and reduced treatment response; trauma-informed interventions with substance use components show promising results.

Stratified systematic review (k=62 studies, 2000–2024) on the co-occurrence of adverse childhood experiences (ACE) and cannabis use in the psychosis spectrum. Significant interactions: OR up to 20,9 in community samples, OR up to 31,0 in first-episode psychosis. Combined exposure associated with earlier disease onset (2,9–3,6 years earlier), more severe positive symptoms, reduced treatment response, poorer functional outcomes. Trauma-informed interventions with an integrated addiction component show promising results.

Synthetic cannabinoids were associated with longer psychosis-related hospital stays than conventional cannabis, as well as with psychotic episodes lasting more than five months in about 30 percent of cases.

Systematic review (PRISMA 2020) on psychiatric consequences of synthetic cannabinoids and novel psychoactive substances across k=58 studies; 28 studies (48.3%) with ≥6-month follow-up, 22 studies (37.9%) rated high quality. Hospitalizations for synthetic cannabinoid-associated psychosis were significantly longer than for cannabis psychosis (43.45 vs. 22.91 days). In one case series, prolonged psychotic episodes >5 months occurred in ~30% of patients. Synthetic cathinones were associated with a rhabdomyolysis rate of 63%; NBOMe substances with a fatality rate of ~15% and an ICU admission rate of ~40%.

Cannabis use independently increases the risk of psychosis and schizophrenia by approximately threefold (OR 2,9).

SR + meta-analysis (k=7 studies) on cannabis as an independent risk factor for psychosis/schizophrenia; pooled OR=2,9 (95% CI 2,4–3,6), no publication bias or heterogeneity. Dose-dependent effect on psychotic symptoms; vulnerable groups: adolescents, persons with a history of psychotic symptoms, high genetic risk.

Systematic review describes high comorbidity rates between cannabis use and psychiatric disorders in Indian studies, without measuring a specific intervention effect.

Systematic review of Indian studies on comorbid cannabis use and psychiatric disorders; k=52 included articles. n=16 studies described acute psychotic episodes with predominantly positive symptoms following cannabis use; n=6 studies observed increased prevalence of all psychiatric disorders due to cannabis; n=14 studies reported high substance use rates in psychiatric patients.

CBD 600 mg significantly inhibited THC-induced paranoia and hippocampus-dependent memory impairment compared with placebo.

n=48 healthy subjects (CBD n=22, placebo n=26), oral CBD 600 mg vs. placebo before intravenous THC (1,5 mg). Clinically significant positive psychotic symptoms (PANSS ≥3 point increase) less frequent in the CBD group: OR=0,22 (χ²=4,74, p<0,05). Paranoia (SSPS) lower in the CBD group (t=2,28, p<0,05). Episodic memory (HVLT-R): -0,4% CBD vs. -10,6% placebo (t=2,39, p<0,05). Demonstrates a protective effect of CBD against THC-induced psychosis risk symptoms.

THC induces psychotomimetic symptoms and memory deficits; CBD does not attenuate these effects in a 2:1 ratio, but shows an antipsychotic effect in isolation in light users.

Randomized crossover RCT (n=48 cannabis users, stratified by schizotypy and frequency of use); inhaled THC (8 mg) significantly increased psychotomimetic scores (PSI) and negative symptoms (BPRS) as well as impaired episodic and working memory. CBD alone reduced PSI scores only in light users. CBD (16 mg) in a 2:1 ratio did NOT attenuate THC-induced psychotomimetic effects and memory impairment.

Frequent cannabis users show an attenuated response to the psychotomimetic, amnestic and endocrine effects of i.v. THC, but not to its euphoric effect.

RCT (n=52: 30 frequent users, 22 healthy controls), IV Δ9-THC 0/2.5/5 mg vs. placebo. THC induced dose-dependent psychotomimetic effects and memory impairments in both groups. Frequent users showed significantly attenuated responses to psychotomimetic, cognitive and anxiogenic THC effects (development of tolerance), but no attenuation of the euphoric effect. Finding demonstrates tolerance to psychotomimetic cannabis effects with frequent use, not their absence.

CBD did not reduce the acute harmful effects of THC on memory, psychosis or other parameters at any dose tested.

n=46 healthy, occasional cannabis users; RCT with 4 CBD:THC ratios (0:1, 1:1, 2:1, 3:1). THC (10 mg) induced positive psychotic symptoms on the PANSS (d=0.69, p<0.00001); no CBD dose (10/20/30 mg) significantly modulated this effect. No protection by CBD against acute psychotomimetic THC effects at clinically relevant CBD:THC ratios.

Δ⁹-THC dose-dependently reduced P300a and P300b amplitudes, without affecting latencies or early sensory components.

Randomized cross-over trial (n=26), intravenous Δ⁹-THC (placebo, 0,015, 0,03 mg/kg): THC reduced P300a and P300b amplitudes dose-dependently (p<0.05 each) and dose-dependently induced psychotomimetic effects as well as perceptual alterations; the THC-induced P3b reductions correlated with the perceptual alterations. Findings support CB1 receptor involvement in top-down/bottom-up information processing as a neurobiological substrate of cannabis-associated psychosis vulnerability.

Δ9-THC dose-dependently increases cortical neural noise, which correlates with psychosomimetic positive and disorganization symptoms.

n=24 healthy volunteers, randomized crossover RCT with i.v. Δ⁹-THC (placebo, 0.015, 0.03 mg/kg); THC increased neural noise (Lempel-Ziv complexity) dose-dependently; strong positive correlation between neural noise and psychosis-like positive and disorganization symptoms (independent of signal power); negative-symptom-like effects did not correlate with noise.

Intravenous THC induced transient positive psychosis symptoms as well as anxiety and cognitive impairments compared to placebo.

n=22 healthy adult men (age 28±6 years), IV THC 2,5 mg vs. placebo (double-blind); THC induced positive psychotic symptoms (PANSS, CAPE: significantly increased vs. placebo), increased anxiety ratings, worsened neuropsychological performance (RAVLT, Digit Span, Verbal Fluency); no association between psychotic symptoms and anxiety intensity (r not significant) or working memory deficits — THC-induced transient psychosis is independent of anxiety and cognitive impairment.

Δ9-THC transiently exacerbated core cognitive and psychotic symptoms in schizophrenia patients.

n=13 stable antipsychotically treated schizophrenia patients, i.v. Δ9-THC 0/2,5/5 mg, double-blind, placebo-controlled; THC transiently increased positive, negative and general symptoms (PANSS), learning/memory deficits, perceptual disturbances, akathisia as well as prolactin and cortisol. Schizophrenia patients responded more vulnerably to cognitive THC effects than healthy control subjects.

THC worsened inhibitory control and correlated with the extent of acute psychotic symptoms.

n=36 healthy men; acute oral THC (10 mg) vs. placebo (double-blind, crossover); THC increased inhibition errors and attenuated left frontal activation; severity of psychotic symptoms correlated directly with inhibition error frequency (p<0,05) and inversely with inhibition efficiency — experimental evidence for a neural mechanism of THC-induced psychotic symptoms.

THC induced pronounced acute psychological and physiological effects, CBD was indistinguishable from placebo.

Randomised double-blind cross-over trial n=16 healthy male subjects; oral THC 10 mg vs. CBD 600 mg vs. placebo; THC vs. placebo: significant positive psychotic symptoms, anxiety, dysphoria (AUC and effect at 2h: p<0,01); CBD: no significant differences from placebo on psychotic or anxiety symptoms — CBD safety signal with regard to psychosis.

Substance abuse, particularly of cannabis and alcohol, is associated with a significantly increased risk of schizophrenia.

Nationwide Danish registry study (n=3.133.968 individuals, 105.178.673 person-years): cannabis abuse diagnosis increased the risk of developing schizophrenia to HR=5,20 (95% CI 4,86–5,57); strongest association of all substances (alcohol HR=3,38). Risk remained significant 10–15 years after substance abuse diagnosis; adjustment for sex, urbanicity, co-abuse, psychiatric history, parental substance abuse and socioeconomic status.

High cannabis use increases the risk of schizophrenia sixfold, independent of other psychiatric disorders and social background.

Landmark cohort study (n=45.570 Swedish conscripts, 15 years follow-up); high users (>50 occasions) RR=6,0 (95% CI 4,0–8,9) for schizophrenia; dose-response gradient; cannabis confirmed as an independent schizophrenia risk factor after multivariate adjustment.

Daily use of high-potency cannabis markedly increases the risk of first-episode psychosis and explains a substantial part of the site-related incidence differences.

Multicentre EU-GEI case-control study (n=901 cases, n=1237 controls, 11 European sites): daily cannabis use OR=3,2 (95% CI 2,2–4,1) for psychotic disorder; daily high-potency cannabis OR=4,8 (95% CI 2,5–6,3). Population attributable fraction (PAF) for high-potency cannabis: 12,2% overall, 30,3% in London, 50,3% in Amsterdam. Incidence rate correlated positively with high-potency prevalence (r=0,7; p=0,0286) and daily use (r=0,8; p=0,0109).

Cannabis use is dose-dependently associated with significantly increased odds of psychosis and schizotypal personality disorder.

NESARC study (n=34.653, nationally representative US sample, 2004–2005); cannabis use dose-dependently associated with psychosis risk: lifetime use OR=1,27 (95% CI 1,03–1,57), cannabis abuse OR=1,79 (95% CI 1,35–2,38), cannabis dependence OR=3,69 (95% CI 2,49–5,47); schizotypal personality disorder also dose-dependently increased (lifetime use OR=2,02; dependence OR=7,32, 95% CI 5,51–9,72).

Cannabis use is associated with significantly increased prodromal psychosis symptoms in adolescents.

Population-based study (n=6.330 adolescents, 15–16 years); cannabis use associated with prodromal psychosis symptoms (OR=2,23; 95%-CI 1,70–2,94), adjusted for covariates; dose-response effect; to date the largest unselected population-based study on this topic.

Dose-dependent increased risk of schizophrenia with cannabis use, not explainable by other drugs or personality traits.

n=50.087 Swedish conscripts (1969 cohort), long-term follow-up for schizophrenia hospitalisation. Dose-dependent increase in risk: adjusted OR for >50 times use 6,7 (95% CI 2,1–21,7) in the cannabis-only group; OR trend for increasing use 1,2 (95% CI 1,1–1,4, p<0,001). Association robust after controlling for other drugs and personality traits.

Frequent cannabis use in adolescence significantly increases the risk of psychosis even after controlling for relevant confounders (HR=3,0).

Prospective cohort (n=6.534, North Finland birth cohort 1986); adolescent cannabis use (≥5 times) associated with increased psychosis risk up to age 30: HR=6,5 (95% CI 3,0–13,9); after adjustment for prodromal symptoms, parental psychosis and other substance use HR=3,0 (95% CI 1,1–8,0), p<0,05.

Cannabis use increases the risk of psychosis incidence and considerably worsens the prognosis in existing vulnerability to psychosis.

3-year follow-up, n=4.045 psychosis-free individuals (NL general population); baseline cannabis use increased the risk of psychotic symptoms (OR=2.76, 95% CI: 1.18–6.47), of severe psychotic symptoms (OR=24.17, 95% CI: 5.44–107.46) and of need for clinical care (OR=12.01, 95% CI: 2.24–64.34). Over 50% of psychosis diagnoses over the course of the cohort attributable to cannabis use.

Cannabis use significantly increases the risk of psychotic symptoms, particularly pronounced in persons with psychosis predisposition.

Prospective cohort study (n=2437, age 14–24 years), 4-year follow-up: Cannabis use increased the cumulative occurrence of psychotic symptoms (aOR=1.67, 95% CI 1.13–2.46). In persons with pre-existing psychosis predisposition, the risk was markedly more pronounced (adjusted risk difference 23.8%, 95% CI 7.9–39.7, p=0.003); in persons without predisposition, also significant (5.6%, p=0.033). Dose-response relationship with increasing frequency of use demonstrated.

Continued cannabis use is associated with increased relapse risk for psychosis (dose-dependent, OR 1,13 for cannabis use vs. no use; OR 1,07 for pattern change).

n=1.055 patients with first-episode psychosis over 6 years; continuous cannabis use after first manifestation of psychosis associated with 4-fold increased relapse risk (HR=3.89, 95%CI 2.64-5.73, p0.001); dose-dependent effect with strongest association for daily use.

Adolescent cannabis use significantly increases the risk of subclinical psychosis symptoms in adulthood.

Prospective community cohort (n=591, observation period 1978–2008, 30 years); regular cannabis use in adolescence associated with schizotypal symptoms in adulthood (OR=2.60, 95% CI 1.59–4.23, multivariate); bivariate analysis: OR=2.29 (95% CI 1.32–3.97). Polydrug use (including cannabis) associated with core schizophrenia symptoms (OR=1.75, 95% CI 1.01–3.03).

High-potency cannabis is significantly more frequent in first-episode psychosis patients than in healthy controls, indicating an increased psychosis-related risk.

Case-control study, n=280 first-episode psychosis vs. n=174 healthy controls; daily cannabis use OR=6,4 vs. controls; use >5 Jahre OR=2,1; among cannabis users: 78% of cases vs. 37% of controls used high-potency cannabis (sinsemilla/skunk), OR=6,8 — delta-9-THC potency as a significant risk factor.

Cannabis-induced attenuated psychotic symptoms are a strong predictor of transition to manifest psychosis in UHR individuals.

Prospective cohort study, n=190 UHR individuals (ultra-high-risk for psychosis), mean follow-up 5,0 years. Individuals with a history of cannabis-induced attenuated psychotic symptomatology (APS) had a 4,90-fold increased risk of transition to a psychotic disorder (OR=4,90; 95% CI 1,93–12,44; p=0,001). Higher severity of cannabis abuse also predicted psychosis transition (p=0,036), mediated via higher abuse in the APS subgroup.

Frequent cannabis use in adolescence increases the long-term risk of schizophrenia approximately 3,7-fold compared to non-users.

Prospective cohort study (n=50.087 recruits, 35 years follow-up); frequent cannabis use in adolescence vs. no use: OR=3,7 (95%-CI 2,3–5,8) for schizophrenia, OR=2,2 (95%-CI 1,0–4,7) for brief psychosis; decline in risk over decades in moderate users, but markedly less in frequent users.

Prior cannabis use is associated with a markedly more severe inpatient course of schizophrenia (more readmissions and longer hospitalization).

n=50.087 Swedish men, 357 schizophrenia cases identified; patients with a history of cannabis use had significantly higher disease burden: median duration of first admission 59 vs. 30 days, readmission rate 10 vs. 4 times, total hospital days 547 vs. 184; OR for >20 readmissions = 3,1 (95% CI 1,3–7,3), OR for admission >2 years = 2,4 (95% CI 1,1–7,4), after adjustment for confounders.

The covariation between cannabis use and psychotic experiences is predominantly due to shared environmental factors and not genetic influences.

Twin cohort study (n=4830 pairs, 16-year-olds) on cannabis use and psychotic experiences (PE). Cannabis use explains 2-5% of the variance in positive, cognitive and negative PE. Covariation between cannabis use and paranoia, cognitive disorganization as well as negative symptoms fully explained by environmental factors (bivariate CE=69-100%); no evidence of a direct causal effect independent of shared environmental risk factors.

Cannabis-induced psychosis is associated with the highest conversion rate to schizophrenia or bipolar disorder (47,4%).

n=6.788 patients with substance-induced psychosis (Danish registry 1994–2014); 32.2% (95%-CI 29.7–34.9%) converted to schizophrenia spectrum or bipolar disorder. Highest conversion rate for cannabis-induced psychosis: 47.4% (95%-CI 42.7–52.3%). Half of all conversions to schizophrenia within 3.1 years; young age and self-harm as significant risk predictors. Published in American Journal of Psychiatry.

Child abuse and cannabis use interact synergistically and increase the risk for positive psychotic symptoms significantly more than either factor alone.

Replication in 2 independent population samples (Greek perinatal cohort n=1.636; NEMESIS n=4.842): cannabis use and childhood traumatization interacted synergistically on psychosis risk [Greece: F(2,1627)=4.18, p=0.02; NEMESIS: χ²(3)=8.08, p=0.04]; psychosis-inducing cannabis effects stronger with prior sexual/physical abuse.

Cannabis use in adolescence robustly increases the risk of psychotic experiences, tobacco use alone shows no significant association after adjustment.

ALSPAC cohort (n=5.300, JAMA Psychiatry): Early cannabis use in adolescence associated with increased risk of psychotic experiences at age 18 after confounder adjustment (OR=3,70, 95% CI 1,66–8,25); late-onset use also significant (OR=2,97, 95% CI 1,63–5,40); tobacco effects attenuated after adjustment (adjusted OR 1,78, 95% CI 0,54–5,88 ns), cannabis effects remained robust.

Longer cannabis use is associated with a significantly increased risk for psychosis-associated outcomes, also after reduction of confounding through sibling-pair analysis.

Prospective birth cohort study (n=3.801, Brisbane) with sibling-pair analysis to control for confounding. Cannabis use ≥6 years associated with: non-affective psychosis (aOR=2.2, 95%-CI=1.1-4.5), hallucinations (aOR=2.8, 95%-CI=1.9-4.1), highest quartile of Peters Delusions Inventory (aOR=4.2, 95%-CI=4.2-5.8). Within sibling pairs, the association between earlier cannabis use and delusional experiences remained significant — reducing the likelihood of confounding by unmeasured variables.

Cannabis increases psychosis risk independently of COMT genotype; a specific gene-environment interaction was not demonstrated.

n=2.630 ALSPAC birth cohort (longitudinal); cannabis use at age 14 → psychotic experiences at age 16. No moderating effect of COMT genotype (no evidence of interaction under multiplicative models). Conclusion: cannabis increases psychosis risk independently of genotype — the warning applies to society as a whole, not only to those genetically predisposed.

Cannabis use significantly increases the risk of incidence and persistence of subclinical psychotic symptoms.

Prospective population cohort study (n=1.923, age 14-24 at baseline, Germany, EDSP study, 10-year follow-up); newly occurring cannabis use increased the risk of later psychotic symptoms (adjusted OR=1,9, 95% CI 1,1–3,1; p=0,021); continued use increased the risk of persistent psychotic symptoms (OR=2,2, 95% CI 1,2–4,2; p=0,016); incidence rate of psychotic symptoms 31% in exposed vs. 20% in non-exposed (baseline→T2), 14% vs. 8% (T2→T3).

CHR individuals consume more cannabis and tobacco than controls, however substance use at baseline does not predict transition to psychosis.

Longitudinal cohort study: n=735 persons with clinical high risk of psychosis (CHR) + n=278 controls, follow-up at 6 and 12 months. CHR participants showed significantly higher cannabis use (severity and frequency) as well as earlier first use. Crucially: baseline cannabis use did NOT differentiate between those who later converted to psychosis (n=90) and non-converters (n=272) — no statistically significant predictive value for transition.

Regular cannabis use in adolescence cumulatively increases the risk of subclinical psychotic symptoms, including paranoia and hallucinations, with a persisting effect after abstinence.

Longitudinal study (n=1.009 boys, age 13–18), regular cannabis use → subclinical psychotic symptoms +21% per year of use (fixed-effects); paranoia odds +133% (p<0.05), hallucination odds +92% (p<0.05); effect persisted after 1-year abstinence. No evidence of reverse causation.

High-potency cannabis use (skunk) is associated with a significantly increased risk of a first episode of psychosis, particularly with daily use.

Case-control study (n=410 first-episode psychosis patients, n=370 controls), South London: high-potency cannabis (skunk) OR=2,92 (95%-CI 1,52–3,45; p=0,001) for psychotic disorder; daily skunk use OR=5,4 (95%-CI 2,81–11,31; p=0,002). Population attributable fraction for skunk use in South London: 24% (95%-CI 17–31%) of all first-episode psychosis cases. 53% of patients used high-potency cannabis.

Persistent cannabis use significantly worsens symptomatology, global functioning and relapse frequency in patients with non-affective psychosis.

n=678 patients with non-affective psychosis (GROUP study, Netherlands/Belgium), 3-year follow-up; persistent cannabis users had significantly more positive symptoms (PANSS positive p<0,001), more general symptoms (PANSS general p<0,001), worse global functioning (GAF symptoms p=0,017; GAF disability p<0,001) and more frequent psychosis relapses (p=0,038) compared to non-users and those who discontinued use.

Only very early cannabis use (≤14 years) is associated with an earlier onset of psychosis; CUD overall shows no significant effect on age of onset of illness.

Retrospective analysis of n=625 first-episode psychosis patients (age 14–29 years); 75% had a cannabis use disorder (CUD); cannabis onset before psychosis onset in 87,6%. Early CUD onset (≤14 years) vs. non-CUD: younger age of onset of illness (F(1)=5,2; p=0,024; partial η²=0,026). Earlier age of cannabis onset predicted earlier age of psychosis onset (β=−0,49; R²-change=0,25; p<0,001). Effect size small; CUD without further SUD showed no significant AAO difference vs. NCUD.

Prenatal cannabis exposure and early age of onset increase the risk of psychotic symptoms, whereby the mediation pathway via EAOM was not significant.

Prospective cohort study (n=596 offspring, follow-up to age 22 years): both prenatal cannabis exposure (PME) and early age of onset (EAOM) significantly predicted increased rates of psychotic symptoms (PS) (total effect PME+EAOM p<0,05). Direct effect of PME on PS marginally significant (p=0,06) after controlling for EAOM and further covariates.

Cannabis-induced psychosis is associated with a high risk of later schizophrenia-spectrum disorders and earlier first manifestation.

Prospective follow-up cohort (n=535) after cannabis-induced psychosis (Danish Psychiatric Register, 1994–1999, follow-up ≥3 years): 44,5% developed schizophrenia-spectrum disorders; 77,2% experienced new psychotic episodes of any kind. Schizophrenia occurred earlier than in the comparison group (men 24,6 vs. 30,7 years). Risk factors: male sex and young age.

Cannabis reduction/cessation after treatment initiation was associated with the strongest symptom improvement, while continued use was associated with persistent symptomatology.

Longitudinal cohort study (n=502 first-episode psychosis patients, 7 London FEP teams, 2 assessment time points): Cannabis use associated with younger age at first presentation and manic symptoms as well as conceptual disorganization (not with hallucinations or negative symptoms). Patients who reduced/stopped cannabis use after first contact showed >the greatest symptom improvement after 1 year compared to continued users; continued users remained more symptomatically burdened than never-users.

Daily use of high-potency cannabis shifts psychosis onset forward by an average of 6 years.

n=410 first-episode psychosis patients, Cox proportional hazards model: Cannabis use → earlier first onset of psychosis HR=1,42 (95%-CI 1,16–1,74; p<0,001). Early onset ≤15 years HR=1,40 (95%-CI 1,06–1,84; p=0,050). Daily high-potency cannabis (skunk) → earliest first manifestation: mean age 25,2 years vs. 31,4 years in never-users (HR=1,99; 95%-CI 1,50–2,65; p<0,0001); first manifestation on average 6 years earlier than in non-users.

No influence of cannabis use on positive symptomatology, but slightly worse functioning with higher use.

Prospective longitudinal study (n=160 cannabis users vs. n=167 other substance users, 24 months); cannabis dose NOT associated with severity of positive symptoms, and no symptom improvement even after abstinence. However, higher cannabis exposure associated with worse psychosocial functioning level (small effect size). No dose effect on positive symptomatology demonstrated.

Continuous cannabis use was associated with persistently higher psychotic symptomatology and worse functioning; cessation of use significantly reduced symptomatology.

n=314 first-manifestation psychosis patients (OPUS trial), 5-year follow-up; continuous cannabis use associated with higher psychosis dimension scores [difference 0,97; 95%-CI 0,41–1,53; p<0,001] and lower GAF [difference 8,26; 95%-CI 2,13–14,39; p=0,01]; cessation of cannabis between baseline and 5-year follow-up significantly fewer psychotic symptoms [adjusted difference -1,04; 95%-CI -1,77 to -0,31; p=0,006].

Continued daily use of high-potency cannabis after first psychotic event significantly increases relapse risk compared to abstinence.

n=256 patients with first manifestation of psychosis, prospective follow-up over 2 years; continuous high-frequency users of high-potency cannabis (skunk) had a significantly increased relapse risk compared to abstinent patients (OR 3,28; 95% CI 1,22–9,18), more relapse episodes (IRR 1,77) and more intensive psychiatric care (OR 3,16; 95% CI 1,26–8,09).

Continued cannabis use after first-episode psychosis increases relapse risk, partially mediated by poorer medication adherence.

Prospective cohort n=245 patients with first-episode psychosis (2-year follow-up); 37% experienced relapse; ongoing cannabis use predicted relapse; medication adherence mediated 20–36% of the effect (β_indirect=0,08; 95% CI 0,004–0,16 for relapse risk); model explained R²=0,25 of relapse variance.

High-potency cannabis use after psychosis onset significantly increases the risk of medication non-adherence compared to non-users.

Prospective cohort study (n=233) in patients after first-episode psychosis; continued use of high-potency cannabis forms (skunk-like) associated with OR=5.26 [95% CI 1.91–15.68] for medication non-adherence compared to non-users (51% vs. 83% non-adherence; p<0.05); no significant effect with sporadic or mild-potency cannabis use.

Cannabis use is bidirectionally associated with a less favorable course of psychotic symptoms and with earlier psychosis onset.

n=229 patients with schizophrenia spectrum disorder, 10-year follow-up after first admission; lifetime prevalence of cannabis use 66,2%; cannabis use bidirectionally associated with psychotic symptoms (mixed-effects model, p<0.05 after adjustment for sex, age, SES, other substances, antipsychotics); lifetime use associated with earlier psychosis onset.

Persistent cannabis use worsens symptomatology and functional level in psychosis; cessation of use normalises outcome to the level of never-users.

10-year cohort study (n=209 first-episode psychosis patients, PAFIP cohort): Persistent cannabis users showed significantly more severe symptoms (BPRS: χ²=15.583, p<0.001; SAPS: χ²=12.386, p=0.002) and worse functionality (DAS: χ²=6.067, p=0.048; GAF: χ²=6.635, p=0.033) compared to ex-users and never-users. Ex-users showed outcomes similar to never-users.

Cannabis-Use-Disorder as well as stimulant abuse worsen functional and symptomatic outcomes in first-episode psychosis patients over 2 years.

2-year prospective cohort (FEP patients, 18–30 years, Montreal); Cannabis-Use-Disorder (CUD) alone was the only group with symptomatic and functional worsening from year 1 to year 2. Worse 2-year outcomes in the CUD group vs. no-SUD group; persistent CUD without polysubstance identified as an independent risk factor for course worsening in psychosis.

Involuntary first admission and cannabis abstinence are associated with improved long-term treatment adherence in FEP, however adherence remains persistently poor in a large proportion.

8-year longitudinal study (N=98 FEP patients); 52% cannabis use at baseline, 74,4% poor treatment adherence. Cannabis abstinence during follow-up was associated with improved adherence in 38,9% of patients; involuntary first admission as a further predictor (p<0,001). Continued cannabis use (persistent CUD) associated with persistently poor treatment adherence over 8 years — intervention target for FEP treatment.

Cannabis abuse significantly increases the risk of psychotic relapses in schizophrenia, more pronounced dose-dependently with severe abuse.

Prospective 1-year cohort study (n=93: 24 cannabis users vs. 69 non-users) in patients with early-onset schizophrenia. Significantly more and earlier psychotic relapses in the cannabis group (p=0,03); association stronger with severe use (p=0,002). No confounding by other drugs. In all but one patient, cannabis use preceded the first psychotic symptom by >1 year.

Discontinuing cannabis after first psychosis improves functional long-term outcome and reduces negative symptoms, while continued use is associated with a worse course.

Prospective 8-year cohort (n=92) after first psychosis episode; patients who stopped cannabis (CUS, n=27) showed better functional long-term outcome vs. continuing users (CU, n=25): effect size=1.26 (95% CI 0.65-1.86) and fewer negative symptoms (effect size=-0.72, 95% CI -1.27 to -0.14). Cannabis cessation after first psychosis → clearly improved long-term course.

High-potency and daily cannabis use increases the risk of psychosis partly independently of childhood trauma, but without significant synergistic interaction.

Case-control study (n=231 first-hospitalised with psychosis + 214 controls); combination of lifetime cannabis use + childhood abuse increased odds of psychosis by almost threefold (adjusted OR=2,94; 95% CI 1,44–6,02; p=0,003); no significant interaction effect (OR=1,46; 95% CI -0,54–3,46; p=0,152). High-potency cannabis (daily) increased risk of psychosis independently of childhood traumatisation.

Lifetime cannabis use did not increase the rate of psychosis conversion in clinical high-risk subjects and was associated with better social functioning.

Prospective cohort of clinical high-risk adolescents (CHR+, n=101; healthy controls, n=59), mean follow-up 2,97 years; logistic regression: lifetime cannabis use NOT associated with psychosis conversion (n=15 conversions); cannabis users had higher social functioning (GF:Social, p<0.001) compared to non-users.

Cannabis use is significantly associated with an earlier first onset of non-affective psychosis, independent of other substances (HR 2,66).

n=112 patients with first non-affective psychosis; history of cannabis use significantly increased the risk of psychosis onset (HR 2.66, 95% CI 1.74–4.05, p<0.001); median psychosis onset in cannabis users 23.5 vs. 33.5 years in non-users (p<0.001); dose-response pattern confirmed.

The combination of childhood sexual trauma and early cannabis use significantly increases the risk of psychosis multiplicatively compared to the individual factors alone.

NCS community sample; interaction cannabis use <16 years × childhood trauma → clinically diagnosed psychosis: OR=6.93 (95%-CI 1.39–34.63, p=.02). Subgroup cannabis <16 y. with sexual trauma: OR=11.96 (95%-CI 2.10–68.22, p=.01). No significant main effect for cannabis alone; risk occurs primarily in the vulnerable adolescent subgroup.

After adjustment for childhood trauma, cannabis use was no longer significantly associated with psychosis; childhood sexual abuse proved to be a considerably stronger predictor.

Population survey (n=7.403, Adult Psychiatric Morbidity Survey 2007); cannabis use alone not significant for psychosis diagnosis after adjustment for childhood trauma; childhood sexual trauma: OR 6,10 for psychosis; combined trauma + cannabis use: OR 7,84 — cannabis-psychosis association moderated by childhood trauma.

Higher schizophrenia risk allele burden is associated with increased cannabis use, suggesting a shared genetic basis.

n=2082 healthy adults; higher burden of schizophrenia risk alleles was significantly associated with cannabis use (ever vs. never use: p=2,6×10⁻⁴; amount of use among users: p=3,0×10⁻³); finding suggests a shared genetic aetiology between schizophrenia risk and cannabis use.

The association between cannabis use and psychotic experiences was considerably attenuated by adjustment for tobacco use and other drug use and lost statistical significance.

ALSPAC cohort (n=1.756 adolescents): Cannabis use at age 16 years was associated with an increased risk of psychotic experiences at age 18 (OR=1,48, 95% CI 1,18–1,86); after adjustment for tobacco smoking the association attenuated (OR=1,27, 95% CI 0,91–1,76); almost all subjects used cannabis together with tobacco, which complicates separability of the effects.

Cannabis use is associated with psychotic symptoms; 69–84% of this association is explained by shared genetic factors, but there are also individual causal effects within discordant sibling pairs.

n=1.120 twin pairs from the ABCD study; cannabis use associated with psychotic symptoms, but twin analysis shows: association primarily explained by shared genetic vulnerability (polygenic risk score for schizophrenia), not primarily a causal effect of cannabis itself.

Cannabis use in early adolescence is associated with reduced cortical thickness in male adolescents with high polygenic risk score, suggesting an association between cannabis use and genetic schizophrenia risk in brain maturation.

n=1.030 adolescents, early cannabis onset associated with thinner cortical brain maturation in individuals with high polygenic schizophrenia risk score (p0.01), no association in low PRS, supports gene×environment interaction in psychosis risk.

Cannabis abuse predicted psychosis conversion, however this effect was confounded by alcohol use and lost strength after adjustment.

Multisite cohort N=370 clinically high-risk subjects (CHR), mean age 18,3 years; cannabis dependence/abuse (CA/CD, n=67) vs. no use: significantly higher psychosis conversion rate (p=0.031) and vs. non-impaired use (p=0.027); alcohol adjustment attenuated this association — alcohol identified as a confounder.

Earlier cannabis use is significantly associated with earlier onset of high-risk and prodromal symptoms for psychosis.

Prospective multicenter study (n=245 high-risk individuals, 18 months follow-up): Younger age of onset of cannabis use or cannabis use disorder correlated significantly with younger symptom onset for 6 prodromal symptoms (0,33 < r_s < 0,83; p<0,004 to p<0,001). Cannabis onset preceded symptom onset in most participants. Result supports a causal role of early use in the development of psychosis in vulnerable individuals.

Persistent SUD is associated with worse functional and symptomatic courses compared with remitting SUD in FEP patients.

n=212 patients with first-episode psychosis (18–30 y), 2-year follow-up; cannabis was the most frequently abused substance (42.9% at baseline). Persistent substance use disorders (SUD) were associated with worse functional outcomes (QoL, SOFAS, employment), more symptoms (PANSS) and higher use of emergency/hospital services; SUD rate declined by ~30% in the first year. SUD persistence was more predictive than SUD presence at admission.

The combination of cannabis use and childhood trauma increases the risk of psychotic symptoms in adolescence beyond the additive measure of both individual risks.

n=211 adolescents (12–15 years); cannabis use significantly associated with psychotic symptoms; interaction with childhood trauma: greater-than-additive effect (risk from combination > sum of individual risks). Important subgroup moderator evidence: highest psychosis risk in traumatized cannabis users in adolescence.

Comorbid CUD in schizophrenia is associated with an enlarged putamen volume, possibly due to polysubstance use.

Cross-sectional study (n=113 schizophrenia patients [CUD=80, NCUD=33] + 84 healthy controls) with MRI volumetry; patients with comorbid cannabis use disorder (CUD) had significantly larger putamen volume vs. NCUD patients; no association between age of onset or frequency of use and gray matter volume; CUD patients more often polysubstance users.

Lifetime cannabis use did not generally increase transition risk, however early onset and frequent use significantly increased the risk.

n=182 ultra-high-risk (UHR) persons for psychosis, 2-year follow-up. Early cannabis onset (<15 years) and frequent use: transition to psychosis 25% vs. 5% with less frequent/later use (χ²=10,971, p=0,001). In the overall sample no significant increase in risk from lifetime use (12,7% vs. 18,8%, p=0,303); risk signal concentrated on early intensive use.

Cannabis and tobacco use did not predict psychosis conversion in CHR individuals; only low alcohol consumption showed a predictive contribution.

Prospective cohort study (n=170 individuals with clinically high risk of psychosis), 4-year follow-up (PREDICT study, 3 sites); Cox proportional hazards analysis: cannabis use at baseline was not a significant predictor of conversion to psychosis (no significant HR), while low alcohol consumption was identified as a predictor.

Cannabis use correlates with stronger positive symptoms in first-episode psychosis, but abstinence is associated with the best symptomatic improvement after 6 months.

n=110 children and adolescents (9–17 years) with first-episode psychosis (FEP); 32 cannabis users (CU) vs. 78 non-users (NCU). CU initially had higher PANSS positive scores (p=0.002) and lower negative (p<0.001) and total scores (p=0.007). After 6 months, CU had significantly lower PANSS scores across all subscales (total p=0.002); former CU who stopped had the best outcome, current CU an intermediate one, NCU the worst. Cannabis use in FEP associated with higher initial positive symptomatology and better course with abstinence from use.

Cannabis use was not associated with an increased rate of psychosis onset in this high-risk group.

n=100 adolescents with ultra-high-risk profile for psychosis; cannabis use/dependence in the year before inclusion was NOT associated with increased risk of psychosis in the following 12 months (32% developed acute psychosis, no significant difference according to cannabis use). Negative finding in a small cohort; methodological limitations (low cannabis use, no monitoring after baseline) limit conclusions.

Continued cannabis use was associated with worse social role functioning, but not with psychopathology or remission status.

Prospective 2-year cohort study (n=124 first-episode psychosis patients, DSM-IV). Continued cannabis use NOT associated with symptomatic or functional remission or clinical recovery. However: cannabis use explained 5,6% variance in economic and 8,4% in social activities (worse social functioning). No significant effect on PANSS positive, negative or general symptoms. Finding: cannabis after first-episode psychosis → modest but measurable deterioration in social functioning; clinical significance moderate.

Only persistent cannabis use, not baseline use alone, is associated with poorer clinical outcomes in adolescent first-episode psychosis.

Representative cohort (n=99) of adolescents (14–18 years) with early-onset first psychotic episode: prevalence of lifetime cannabis use disorder (CUD) 65,7%, current CUD at admission 53,5%, persistent CUD over the course of treatment 26,3%. Only persistent CUD was associated with significantly poorer 18-month outcomes (medium effect sizes after confounder control), while baseline CUD was primarily associated with more severe baseline symptoms and lower psychosocial functioning (p-values not given numerically, but effect sizes medium).

Hopelessness and depression are independent predictors of psychosis symptom severity and partially mediate the relationship between cannabis use and persistent psychotic symptoms; addressing them represents a potentially modifiable treatment target.

Prospective 24-month study on cannabis-induced first-episode psychosis (n=72); hopelessness (beta=0.30, p<0.001) and depression (beta=0.34, p<0.001) independently predict symptom severity. Mediation analysis: hopelessness and depression mediate 31% and 35%, respectively, of the relationship between cannabis use and persistent psychotic symptoms. Combination of high hopelessness, depression and continued cannabis use increases relapse risk (HR=4.32, 95% CI [2.41, 7.75]).

Cannabis abuse was significantly associated with a higher conversion rate to psychosis in high-risk individuals.

n=48 high-risk individuals (CARE program) with subsyndromal psychotic symptoms or familial burden; 1-year follow-up: conversion rate to psychosis in cannabis-dependent/abusing individuals 31,3% vs. 3,1% in non-/minimal users; significant association between cannabis abuse and transition to psychosis.

Chronic cannabis use reduces striatal dopamine synthesis capacity, without this being directly associated with cannabis-induced psychotic symptoms.

PET neuroimaging study (n=19 regular cannabis users vs. n=19 non-users); cannabis use associated with reduced striatal dopamine synthesis capacity (effect size=0,85, p=0,016); inverse correlation between cannabis use frequency and dopamine synthesis (r=-0,77, p<0,001); earlier onset of use associated with greater reduction (r=0,51, p=0,027).

Longitudinal twin study on adolescent cannabis use and adult psychoticism; co-twin control design to control for genetic and environmental confounding. Abstract does not report numerical effect sizes; refers to a consistent association between cannabis use and increased schizophrenia risk (Marconi et al., 2016) as well as a dose-dependent increase in risk with early onset of use (van der Steur et al., 2020).

Cannabis use disorder is associated with increased positive symptoms in psychosis patients, but shows no consistent effect on pre-psychotic symptomatology in UHR subjects.

Cross-sectional study (n=169 psychosis patients + n=59 ultra-high-risk individuals): 45% of psychosis patients and 27% of UHR subjects had a comorbid cannabis disorder. When other substance use disorders were excluded, patients with cannabis disorder showed higher positive symptom scores vs. substance-free control group. Negative correlation between current cannabis use and pre-psychotic negative symptomatology in UHR subjects (direction: more cannabis, fewer negative symptoms — mixed finding).

Cannabis use is associated with more subclinical psychosis symptoms and neuropsychological functional impairments, also in healthy control persons.

n=63 UHR patients (34 cannabis users) + n=58 controls (28 cannabis users), cross-sectional; cannabis-using UHR patients had more basic symptoms than non-using UHR patients; healthy cannabis users achieved verbal memory and fluency performance at UHR level; frequency of cannabis use correlated with severity of several UHR symptoms. Cross-sectional design limits causal conclusions.

Continued cannabis use increases symptom burden in first-episode psychosis, however mediated by lower medication adherence; cannabis users paradoxically showed higher adherence rates.

n=48 first-episode psychosis patients with cannabis abuse (28 continuers, 20 stoppers); after controlling for medication adherence, continuous cannabis users showed significantly higher symptom burden after 12 months (F(1,30)=2,74, p=0,03); adherence higher in continuers (92% vs. 40%, p<0,01), but risk of poorer symptom courses remains elevated.

Cannabis is classified as an independent risk factor for psychoses, while the non-psychoactive cannabidiol is discussed as a possible treatment approach.

Narrative review on the 'cannabinoid hypothesis' of schizophrenia. Cannabis as an independent risk factor for psychosis with a dose-dependent effect and worse disease course. Endocannabinoid system alterations in subgroups (increased CB1 receptor density, increased CSF anandamide levels). Discusses dopamine-endocannabinoid interaction and therapeutic potential of non-psychoactive cannabidiol.