Table 2.
Characteristics of included studies in patients/risk of psychosis comparing cannabis users (P-C) with non-users (P-NC): Neurobiological differences.
| Study | Method | Measure | Population | Cannabis use definition |
N
|
Findings | Limitations | Confounders considered: |
||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| P-C | P-NC | HC-C | HC-NC | ETOH | Other drug | AP | Tob | |||||||
| (a) MRI Studies in high-risk states | ||||||||||||||
| Welch et al. (2011a) | MRI - 2 years longitudinal | Thalamus & mediotemporal volumes | High genetic risk of schizophrenia | Cannabis use during intervening 2 years between scans | 25 | 32 | Bilateral thalamic volume loss P-C vs. P-NC. Amygdala-Hippocampal volume change non-significant | No HC comparator group. Unclear how many progressed to psychotic disorder | Y | Y | N/A | Y | ||
| Welch et al. (2011b) | MRI (cross-sectional) | Volumetric analysis of ventricles, prefrontal lobe, amygdala-hippocampal complex and thalamic nuclei | High genetic risk of schizophrenia | P-C: Regular use; P-NC: Use of ≤3 occasions/lifetime | 73 | 69 | 15 | 20 | Cannabis use correlated with increased ventricular size (particularly 3rd ventricle) in risk group. This relationship does not hold in control group | OR of progression to psychotic disorder 3.18, P-C vs. P-NC. Unclear whether this correlates with imaging findings | Y | Y | N/A | Y |
| Habets et al. (2011) | Discussed below (see (a)(ii)) | |||||||||||||
| Stone et al. (2012) | MRI (cross-sectional) | Regional grey matter (GM) | At-risk mental state (ARMS) | Ever use. Analysis undertaken by cannabis intake (occasions per year) | 19 | 8 | 14 | 13 | Heavier cannabis use associated with reduced GM volume in prefrontal cortex. No distinct effects in ARMs vs. HCs | Small study. Does not establish functional correlate of GM loss | Y | Y | Y | Y |
| Welch et al. (2013) | MRI tensor-based morphometry, 2 years longitudinal | Regional grey matter (GM) | High genetic risk of schizophrenia | Cannabis use during intervening 2 years between scans | 23 | 32 | Reduction in right anterior hippocampus and left superior frontal gyrus P-C vs. P-NC | No HC comparator group. Unclear how many progressed to psychotic disorder | Y | Y | N/A | Y | ||
| Rapp et al. (2013) | Discussed below (see (a)(ii)) | |||||||||||||
| Buchy et al. (2015) | Resting state fMRI functional connectivity (cross-sectional) | Thalamic functional connectivity | Clinical high-risk (NAPLS-2 cohort) | Rated by severity of use, frequency of use, age of first use and whether early or late onset (cut off age 15) | 50 | 112 | 92 | 13 | Significant correlation between age of onset of cannabis use in CHR and thalamic hyper-connectivity with sensory motor cortex (significant on left, trend level on right) | Unclear how many progress to psychotic disorder | Y | N | N | Y |
| Buchy et al. (2016) | MRI (cross-sectional) | Hippocampus, amygdala, thalamus volumes | Clinical high-risk (NAPLS-2 cohort) | Use in the last month | 132 | 387 | 204 | No difference P-C vs. P-NC after adjustment for tobacco and alcohol | No metric for lifetime/cumulative exposure. No HC-C group | Y | N | Y | Y | |
| (b) Structural MRI studies in patients (within 5 years of first episode) | ||||||||||||||
| Cahn et al. (2004) | MRI (cross-sectional) | Regional brain volumes | Recent-onset schizophrenia, schizoaffective, schizophreniform | DSM IV cannabis abuse/dependence | 27 | 20 | No difference total brain volume, caudate, cerebellar volume. Decreased left/right asymmetry P-C vs. P-NC | P-C significantly younger than P-NC (21.13 years vs. 27.61) | N | Y | Y | N | ||
| Szeszko et al. (2007) | MRI (cross-sectional) | Prefrontal brain volumes | Recent-onset schizophrenia, schizoaffective, schizophreniform | DSM IV cannabis abuse/dependence | 20 | 31 | 56 | P-C significantly decreased anterior cingulate grey matter vs. P-NC | Manual outlining of frontal brain regions | Y | Y | Y | N | |
| Wobruck et al. (2009) | MRI (cross-sectional) | Volumetric measure of superior temporal gyrus, amygdala-hippocampal complex, cingulum | Recent-onset schizophrenia, schizoaffective disorder recently admitted to hospital | Lifetime abuse | 20 | 21 | P-C vs. P-NC: no effect on brain morphology between groups | P-C all used cannabis, also reported: stimulants (35%), opiates (10%), cocaine (40%), hallucinogens (5%), alcohol (10%) | N | N | Y | N | ||
| Bangalore et al. (2008) | MRI (cross-sectional) | Dorsolateral prefrontal cortex, hippocampus, posterior cingulate, cerebellum, intracranial volume | Schizophrenia, schizoaffective, schizophreniform | Lifetime use vs. never use | 15 | 24 | 42 | Right posterior cingulate showed trend for reduced GM volume P-C vs. P-NC. No difference for DLPFC, cerebellum, hippocampus, whole brain volume | 8/15 (53%) P-C group had poly-substance use | N | Y | Y | N | |
| Peters et al. (2009) | MRI – Diffusion tensor imaging (cross-sectional) | Fractional anisotropy (FA) & markers of white matter integrity | Recent-onset schizophrenia/schizoaffective/schizophreniform | P-C: Cannabis use < age of 17; P-NC No cannabis use < 17 | 24 | 11 | 21 | FA higher in patients using before 17 than controls in anterior capsule, fasciculus uncinatus, frontal white matter | Small numbers. Requires further replication and correlation with function | Y | Y | Y | N | |
| Dekker et al. (2010) | MRI: High resolution structural and diffusion tensor imaging (cross-sectional) | WM density, fractional anisotropy | DSM IV schizophrenia, male | At least weekly for 6 months of life. Early-Onset Cannabis use ≤15; Late Onset Cannabis Use≥17 | 18 | 8 | 10 | Reduced FA and WM density in left posterior corpus collosum, right occipital lobe, left temporal lobe, for P-NC compared with early-onset users | Small groups size – 8 early-onset users and 10 late-onset users | N | N | Y | N | |
| James et al. (2011) | MRI: voxel-based morphometry and diffusion tensor imaging (cross-sectional) | Regional brain volumes and fractional anisotropy. Neurocognitive IQ (measured by WASI) | Adolescents with DSM IV schizophrenia | More than 3 days/week for 6 months. All participants last use ≥28 days prior | 16 | 16 | 28 | P-C vs. P-NC widespread grey matter loss in mediotemporal, insular, cerebellum, occipital and ventral striatum areas. Decreased FA in brainstem, internal capsule, corona radiata, superior and inferior longitudinal fasciculus | Age of onset younger in P-NC vs. P-C - may represent atypicality of adolescent schizophrenia | N | Y | Y | N | |
| Cohen et al. (2012) | MRI (cross-sectional) | Cerebellar grey and white matter volumes | First-episode schizophrenia | Juvenile cannabis use. Average age of first use (PC-C) 15.1, (HC-C) 15.5; lifetime doses (PC-C) 22700, (HC-C) 17900 | 6 | 13 | 17 | 19 | Dose-related decreased GM in cerebellum for HC-C vs. HC-NC. Decreased GM changes in P vs. HC. No interaction of cannabis use with diagnosis on GM cerebellar changes | Small P-C group | N | Y | N | N |
| Kumra et al. (2012) | MRI (cross-sectional) | Volumes: frontal, temporal, parietal, subcortical, cortical thickness, surface area, cognitive measures | Early-onset schizophrenia | Lifetime cannabis abuse/dependence. Excluded positive urine test | 13 | 35 | 16 | 51 | PC-NC vs. HC-NC and HC-C vs. HC-NC decreased left superior parietal cortex, relative sparing in PC-C. PC-C vs. PC-NS: GM left thalamus volume reduced | Excluded current users | N | N | Y | N |
| Schnell et al. (2012) | MRI (cross-sectional) | Regional brain volumes & neurocognitive testing | DSM IV schizophrenia | Lifetime cannabis abuse/dependence | 30 | 24 | P-C vs. P-NC higher GM density in left middle frontal gyrus. Significantly correlated with Continuous Performance Task (indexes working memory and attention, r=0.681) | Needs further replication | Y | Y | Y | Y | ||
| Cunha et al. (2013) | MRI (cross-sectional) | Grey matter and lateral ventricle volumes & neurocognition (WMAT, COWAT, digit span) | First-episode psychosis | Lifetime history of at least 3 times per month for 1 year | 28 | 78 | 80 | P-NC vs. HC-NC decreased grey matter but in P-C vs. P-NC less grey matter loss in left middle frontal gyrus, left hippocampus, left parahippocampal gyrus. P-C vs. P-NC fewer attentional and executive deficits | P-C younger and more time in full time education than P-NC | Y | Y | Y | N | |
| Haller et al. (2013) | MRI: voxel-based morphometry and diffusion tensor imaging & neurocognitive testing (TAP, WMS-R) (cross-sectional) | Grey matter volume & tract-based spatial statistics & fractional anisotropy | First-episode psychosis | 1. Heavy Use: near daily use for at least 1 year prior to presentation; 2. Light Use>lifetime use 10 times, less than heavy use; 3. Considered as non-user if used up to 10 times | 33 | 17 | No difference between P-C and P-NC on voxel-based morphometry or DTI analysis or neurocognitive measures. No difference between heavy vs. light use | Small size ?underpowered | N | Y | Y | N | ||
| Rapp et al. (2013) | MRI (cross-sectional) | Grey matter volumes in cingulate cortex | At-risk mental state and first-episode psychosis | Current cannabis use | Pts:8 ARMS:14 | Pts:15 ARMS:22 | Negative effect of cannabis use on posterior cingulate cortex and left anterior cingulate for both FEP and ARMS | Small sample size. Manual segmentation – high interrater reliability | Y | N | Y | N | ||
| Malchow et al. (2013) | MRI & MRS (cross-sectional) | Volumetric analysis: hippocampus, amygdala, caudate, putamen, thalamus, corpus callosum; MRS metabolites (N-acetyl-aspartate) indexes neuronal integrity | First-episode schizophrenia | Cannabis abuse | 29 | 20 | 30 | Psychosis patients volume loss left hippocampus and amygdala vs. controls. P-C vs. P-NC larger mid-sagittal area of corpus callosum. P-C had higher left putamen N-acetyl aspartate/choline | No functional correlate. Limited information on cannabis use | N | N | N | N | |
| Epstein and Kumra (2015) | MRI Longitudinal - 18 month follow-up | Cortical thinning & neurocognitive (D-KEFS Tower Test) | Early-onset schizophrenia/schizoaffective/schizophreniform | Lifetime cannabis abuse/dependence at baseline | 11 | 17 | 17 | 34 | No significant main effect for psychosis, but main effect for cannabis use disorder – widespread cortical thinning. No significant effect for cannabis use disorder × psychosis interaction | Small P-C group. Needs replication in a larger cohort | Y | N | N | Y |
| Koenders et al. (2015) | MRI (cross-sectional) | Surface-based analysis of a priori brain regions | DSM IV psychotic disorders (male patients only) | DSM IV cannabis abuse/dependence | 80 | 33 | 84 | P-C vs. P-NC: increased putamen enlargement in P-C. Patients vs. controls: smaller volumes amygdala, putamen, insula, parahippocampus, fusiform gyrus | Did not correct for smoking or medication | Y | Y | Y | Y | |
| (c) Structural MRI Studies in Patients (after 5 years of first episode) | ||||||||||||||
| Rais et al. (2008) | MRI – 5 years longitudinal | Ventricle size and total grey matter (GM) volume change | Recent-onset schizophrenia | Ever cannabis use during scan interval (5 years) | 19 | 32 | 31 | Larger ventricular size and reduced GM in direction: P-C>P-NC>HC. P-C vs. P-NC less pronounced symptomatic improvement | No HC-C group to disentangle psychosis × cannabis interaction | Y | Y | Y | N | |
| Rais et al. (2010) | MRI – 5 years longitudinal | Cortical thickness | Recent-onset schizophrenia | Ever cannabis use during scan interval (5 years) | 19 | 32 | 31 | P-C vs. P-NC no baseline difference. Over 5 years increased cortical thinning in DLPFC, ACC and left occipital lobe | Does not establish functional correlate of cortical thinning | Y | Y | Y | N | |
| Habets et al. (2011) | MRI (cross-sectional) | Cortical thickness | DSM IV psychotic disorders, siblings; controls | 1 Subjects who have never used cannabis; 2 subjects who have used 1–39 times (moderate); 3 subjects who have used ≥40 times (heavy) | 52 pts; 33 siblings | 28 pts; 53 siblings | 21 | 48 | Patients with heavy use had lower cortical thickness than those with no use. Same relationship for siblings but not for controls | Needs replication in larger samples. No function correlate of diminished cortical thickness | N | Y | Y | N |
| Solowij et al. (2011) | MRI (cross-sectional) | Cerebellar grey matter and white matter volume | Schizophrenia, right-handed male | Long-term heavy cannabis use (near daily use for ≥9 years) | 8 | 9 | 15 | 16 | No group differences in GM volume. Direction of WM volume: HC-NC>P-NC>HC-C>P-C | Small groups P-C and P-NC | Y | Y | N | Y |
| Solowij et al. (2013) | MRI (cross-sectional) | Hippocampal shape analysis | Schizophrenia | Long-term regular use; >60000 doses last 10 years | 8 | 9 | 15 | 16 | Hippocampal shape changes in each group vs. HC-NC with greatest changes in P-C vs. H-C | Small P-C group. Chronic patient group. No functional correlation | Y | Y | N | Y |
| Smith et al. (2014) | MRI (cross-sectional) | Surface-based representations of globus pallidus (GP), striatum, thalamus and working memory | DSM IV Schizophrenia | DSM IV lifetime cannabis abuse/dependence | 15 | 28 | 10 | 44 | Morphological shape differences observed in cannabis groups (P-C vs. P-NC and HC-C vs. HC-NC) in striatum, GP and thalamus. Morphological changes more pronounced in P-C than HC-C. For both P-C and HC-C striatal and thalamic changes correlated with WM deficits and younger age of CUD diagnosis | Cannabis use group not had substance misuse diagnosis in 6 months prior to the study. Cross-sectional – need longitudinal follow-up | Y | Y | Y | Y |
| Smith et al. (2015) | MRI (cross-sectional) | Surface-based analysis of hippocampus, episodic memory (logical memory II test) | DSM IV schizophrenia | DSM IV cannabis abuse/dependence 6 months previously | 15 | 28 | 10 | 44 | Effect of shape changes by cannabis use disorder ‘cannabis-like shape’. Separate changes on shape of hippocampus by schizophrenia ‘schizophrenia-like shape’. P-C group demonstrated increasing cannabis shape changes with increasing duration of cannabis use disorder. P-C vs. P-NC trend level worse on episodic memory task | Prolonged period of abstinence for cannabis users (at least 6 months) | Y | Y | Y | Y |
| Rigucci et al. (2015) | MRI – diffusion tensor imaging (cross-sectional) | Fractional anisotropy & markers of white matter integrity | First-episode psychosis (ICD-10 diagnosis confirmed using OPCRIT+) | Lifetime history of cannabis use. Further analyses undertaken on other parameters (age of first use, potency of use, frequency of use) | 37 | 19 | 22 | 21 | Higher potency cannabis associated with disturbed corpus callosum microstructure in both patients with psychosis and cannabis users | No functional correlate described | N | N | N | N |
| (d) functional MRI studies in patients | ||||||||||||||
| Loberg et al. (2012) | fMRI – dichotic auditory perception task | BOLD activation of default mode network and effort mode network | DSM IV schizophrenia | Lifetime cannabis use. Current users excluded | 13 | 13 | P-C vs. P-NC: increased activation in regions involved in effort mode network and decreased activation in default mode network | Unconventional method of indexing network activity. Cannot extend work to current users | N | N | Y | N | ||
| Bourque et al. (2013) | fMRI – emotional memory task | BOLD activation in emotional picture recognition | DSM IV schizophrenia, males | Cannabis use disorder (abuse/dependence) diagnosed in last 6 months | 14 | 14 | 21 | P-C vs. P-NC: medial prefrontal cortex activation increased in emotional picture recognition | Small study. Needs further replication in larger samples | N | Y | Y | N | |
| Potvin et al. (2013) (from same group as Bourque et al (2013)?same patient group) | fMRI – visuospatial task and mental rotation | Regional BOLD activation | DSM IV schizophrenia, males | Cannabis use disorder (abuse/dependence) diagnosed in last 6 months | 14 | 14 | 21 | P-C vs. P-NC: preserved activation left superior parietal gyrus (decreased in P-NC). No difference in task performance P-C vs. P-NC | Small study. Needs further replication in larger samples | N | Y | Y | N | |
| Machielson et al. (2014) | fMRI attentional bias using a classical Stroop task and cannabis Stroop task after 4 weeks of treatment with risperidone or clozapine | Stroop task performance and BOLD activation. Classical Stroop indexes selective attention, Cannabis stroop to index attentional bias | DSM IV schizophrenia, schizoaffective, schizophreniform disorder, male gender | Cannabis abuse/dependence | 28 | 8 | 19 | No difference P-C vs. P-NC in performance in Stroop task. P-C vs. P-NC no difference in regional activation for classical Stroop. Greater activation in left and right amygdala for P-C vs. P-NC | Small P-NC group ?underpowered. Also main purpose of study not to compare P-C vs P-NC groups but to compare risperidone vs. clozapine. Confounders not adjusted for outcomes of interest | N | N | N | N | |
| Peeters et al. (2015) | Resting state fMRI functional connectivity (cross-sectional) | Dorsolateral prefrontal cortex functional connectivity (DLPFC-fc) & neurocognitive testing (WAIS III and others) | Psychosis, unaffected siblings and controls | Ever use vs. never use | Patients with psychosis (n=73); Unaffected siblings (n=83); Controls (n=72); Numbers not clearly stated by cannabis use | No significant group × cannabis interaction for DLFPC-fc and no significant interaction with neurocognitive testing | Most patients on medication. Effect of this unclear. Use of current use maybe more sensitive to change than lifetime cannabis use | N | N | Y | N | |||
| Machielson et al. (2017) | Cue reactivity to cannabis and neutral images fMRI | Regional BOLD activation to cannabis images | DSM IV schizophrenia/schizoaffective/schizophreniform | Lifetime cannabis use disorder (based on CIDI) | 30 | 8 | 20 | P-C vs. P-NC greater activation to cannabis images in right amygdala and left and right thalamus. Also 4 weeks of clozapine superior to risperidone in reducing craving in P-C | Designed to compare clozapine vs. risperidone rather than to compare P-C vs. P-NC. Small P-NC group | N | N | N | N | |
| (e) Other neurobiological measures | ||||||||||||||
| Jockers-Scherubl et al. (2003) | Blood (cross-sectional) | Serum nerve growth factor (NGF) (cross-sectional) | DSM IV schizophrenia, presenting as inpatient, medication naïve | >0.5 g on average per day for at least 2 years. Positive UDS excluded | 21 | 76 | 11 | 61 | P-C NGF levels mean=412.9; P-NC 26.3; HC-C 20.1; HC-NC 33.1. Significantly raised in P-C group | Small numbers in cannabis-using group. P-C onset of psychosis younger than P-NC | Y | Y | Y | N |
| Jockers-Scherubl et al. (2004) | Blood (cross-sectional) | Serum brain derived neurotrophic factor (BDNF) (cross-sectional) | DSM IV schizophrenia, medication naïve | >0.5g on average per day for at least 2 years. Positive UDS excluded | 35 | 102 | 11 | 61 | P-C BDNF levels mean=17.7; P-NC 13.1; HC-C 13.1; HC-NC 13.2. Highest in P-C group, significant vs. H-C and P-NC | P-C onset of psychosis younger than P-NC. Needs replication | Y | Y | Y | N |
| Jockers-Scherubl et al. (2006) | Blood (including follow-up of Jockers-Scherubl et al. (2003)) | Serum nerve growth factor (NGF) (cross-sectional) | DSM IV schizophrenia, treated with medication for psychosis for 4 weeks | >0.5 g on average per day for at least 2 years. Positive UDS excluded | 42 | 66 | 24 | 51 | No statistically significant difference between groups (i.e. NGF levels raised in P-C had normalized). Decrease across patient groups in NGF levels from Jockers-Scherubel et al. (2003). Also small cohort replicated Jockers-Scherubel et al. (2003) baseline results | Needs replication | Y | Y | Y | N |
| Benson et al. (2007) | Eye movement tracking (cross-sectional) | Fixation clustering, saccadic gaze | DSM IV cannabis-induced psychosis (CIP) vs. first-episode schizophrenia based on SCID | Based on diagnosis of CIP (see limitations) | 6 | 11 | 22 | Differences noted in visual scan paths. More restricted visual features in P-C patient with increased fixation clustering such that P-C>P-NC>HC-C. Alterations in saccadic gaze (frequency, amplitude, velocity) for P-C vs. HC-NC in altered pattern compared with P-NC vs. HC-NC | Small study and needs replication. Definition of P-C and P-NC in study based on CIP vs. schizophrenia. Three patients with schizophrenia (in P-NC) had used cannabis before developing schizophrenia | N | N | N | N | |
| Rentzsch et al. (2007) | EEG with auditory stimuli (cross-sectional) | P50 Sensory gating | DSM IV schizophrenia/schizoaffective | History of chronic cannabis abuse. All users abstinent ≥28 days | 15 | 12 | 11 | 18 | No difference between P-C and P-NC. In HC-C P50 sensory deficit correlated with number of years with daily consumption. Relationship not present in other groups | Small sample size ?underpowered. There is a difference of circa 10% between P-C vs. P-NC but not significant | Y | Y | N | N |
| Rentzsch et al. (2011) | EEG with auditory stimuli (cross-sectional) | Mismatch negativity (MMN) | DSM IV schizophrenia/schizoaffective | Chronic cannabis use: at least 5 days per week for at least 1 year by self-report. All users abstinent ≥28 days | 27 | 26 | 32 | 34 | Frequency MMN amplitude P-NC<P-C<HC-C<HC-NC | No functional correlation shown. Requires replication | Y | Y | Y | Y |
| Scholes-Balog and Martin-Iverson (2011) | Electrophysiology with auditory stimuli (cross-sectional) | Prepulse Inhibition (PPI) | Schizophrenia/schizoaffective | Lifetime use | 20 | 44 | 34 | 32 | Alterations of PPI in all groups vs. HC-C such that reduced PPI in P-NC vs. HC-NC but this is diminished in P-C vs. HC-NC. Authors suggest cannabis may have medicating effect on impaired attentional modulation for patients and have similar effect on attentional modulation for patients and healthy controls | Preponderance of other substance abuse/dependence in last 12 months in P-C group (60%) | N | N | Y | Y |
| Pesa et al. (2012) | EEG with auditory stimuli (cross-sectional) | Mismatch Negativity (MMN) and P3a latency | Early psychosis, DSM IV psychotic disorders | History of past or current cannabis use at least monthly for one year | 22 | 22 | 21 | For MMN amplitude: P-NC<P-C<HC-NC. For MMN latency: P-C>P-NC and HC-NC. For P3a amplitude at frontal electrode: P-C<P-NC and latency P-C>P-NC | No functional correlate identified | Y | Y | N | N | |
| Van Tricht et al. (2013) | EEG with auditory oddball paradigm (cross-sectional) | N100, N200, P200, P300 | Ultra-high risk | History of use ≥5 times/lifetime. Use in last month | 19 | 29 | 21 | 29 | No significant differences between P-C and P-NC | Gender differences between groups (18% female P-C; 45% female P-NC). Although tobacco use was adjusted for, other recreational drug use not accounted for | Y | Y | Y | N |
| Cassidy et al. (2014) | Electrophysiology – EEG and facial electromyography to visual stimuli of natural and cannabis rewards (cross-sectional) | Late positive potential (LPP), facial electromyography, skin conductance. Follow-up after 1 month for cannabis usage | DSM IV schizophrenia/schizoaffective | Active cannabis use disorder in past 1 month vs. no cannabis use in last 3 months (P-NC) | 20 | 15 | 20 | 15 | P-C show blunted responses to natural rewards but spared response to cannabis stimulus. LPP in P-C predicts cannabis usage at 1 month | Cannabis images prepared specifically for this task and not validated in previous tasks | N | Y | Y | Y |
| Hagenmuller et al. (2014) | EEG with somatosensory evoked potential (cross-sectional) | N20, P25 | Ultra-high risk (UHR) assessed by structured interview for prodromal symptoms and high risk (HR) for schizophrenia assessed by schizophrenia proneness interview | HR 13; UHR 12 | HR 36; UHR 61 | 3 | 42 | P-C vs. P-NC: in both HR and UHR groups cannabis users showed higher N20-P25 source strength than non-users. Not calculated in HC due to small numbers of HC-C | Small cannabis arms. Possible effect of confounding variables (see right) | N | N | N | N | |
| Winton-Brown et al. (2015) | Electrophysiology with auditory stimuli (cross-sectional) | Prepulse inhibition (PPI) and prepulse facilitation (PPF) | At-risk mental state | Urine drug sample positive | 6 | 18 | 5 | 18 | PPI: HC-C vs. HC-NC increased PPI, P-C vs. P-NC decreased PPI; PPF: No group × substance use interaction | Small numbers of cannabis participants. UDS only indicates recency of use | N | Y | Y | Y |
| Morales-Munoz et al. (2015) | Electrophysiology with auditory stimuli (cross-sectional) | Prepulse inhibition (PPI) | First-episode psychosis, on treatment (assessed by SCID for DSM IV) | Lifetime cannabis abuse/dependence (assessed by SCID), currently abstinent | 21 | 14 | 22 | PPI: at 30 ms reduced PPI for P-C vs. HC-NC and P-NC vs. HC-NC. At 60 ms reduced PPI for P-NC vs. HC-NC but not for P-C vs. HC-NC. At 120 ms no difference between groups | No assessment of alcohol or other drug use. Although all had been on pharmacotherapy cumulative doses not matched | N | N | Y | Y | |
| Rentzsch et al. (2016) (some overlap of participants with Rentzsch et al., 2007) | EEG to auditory novelty and oddball paradigms (cross-sectional) | P300 | Schizophrenia | Chronic cannabis use: at least 5 days per week for at least 1 year by self-report. All users abstinent ≥28 days | 20 | 20 | 20 | 20 | Different effects of cannabis use on patients and controls. Unadjusted: Early novelty P300 HC-NC>HC-C>P-NC>P-C. Late novelty P300 HC-NC>HC-C>PC-NC~–P-C. Parietal oddball HC-NC>HC-C>P-NC>P-C | No significant diagnosis interaction for early novelty p300 and oddball paradigm remained when nicotine and alcohol use entered as covariates. Significant difference between illicit drug use between groups | Y | N | Y | Y |
(a) MRI studies in high-risk groups (genetic, familial, clinical high risk).
(b) MRI structural in patients with psychosis (including volumetric, morphometry and shape analysis, diffusion tensor imaging) first 5 years.
(c) MRI structural in patients with psychosis (including volumetric, morphometry and shape analysis, diffusion tensor imaging) after 5 years.
(d) Functional MRI studies in patients with psychosis.
(e) Other neurobiological measures (EEG, eye-tracking).
P-C: Psychosis/at-risk patients with cannabis use; P-NC: Psychosis/at-risk patients without cannabis use; HC-C: Non-psychosis controls with cannabis use; HC-NC: Non-psychosis controls without cannabis use; ETOH: Alcohol; AP: medication for psychosis; Tob: Tobacco.