Clinical Effects of Cannabis Use in First-Episode Acute Psychotic Patients

Abstract

Introduction:

Cannabis is the most widely used illegal psychoactive substance worldwide. Although there are clinical studies examining the differences between psychotic symptoms emerging after cannabis use and non-cannabis-related psychotic symptoms, data are limited in the scientific literature. We aimed to investigate the effects of cannabis use on psychotic symptoms and compare the cognitive function differences between the cannabis-user and non-user groups.

Methods:

First-episode psychotic patients were included in the study and divided into two groups based on cannabis use. Participants with cannabis use and without cannabis use were compared in terms of socio-demographic factors, psychotic symptoms, cognitive functions, and childhood trauma. All patients were assessed twice, during recruitment and after treatment for psychotic symptoms.

Results:

A total of 38 patients comprising of 18 patients with a history of cannabis use and 20 patients without a history of cannabis use were included. There were significant correlations between cannabis use, and living apart from family and family history of substance abuse. Negative psychotic symptoms were higher in the non-user group. The negative psychotic symptoms improved more significantly after treatment in the non-user group. There was a significant correlation between cannabis use and Stroop Color-Word Test scores.

Conclusion:

Our results give us a chance to argue that psychosis has different features in people with and without a history of cannabis use. These two phenomena could differ in many ways, so different prevention strategies and treatment options should be considered.

INTRODUCTION

Cannabis is the most widely used illicit substance, according to a report published by United Nations Office for Drugs and Crime in 2018. Cannabis was used by 192 million people last year (1), and because of the broad legalization of cannabis in recent years, this usage rate would likely get much higher. It has been shown that the higher frequency and severity of cannabis use increases the risk of developing psychotic symptoms, and especially cannabis use during adolescence is associated with an increased risk for psychosis. Cannabis use is also associated with earlier age of onset for psychosis (2).

Delusions of persecution, depressive symptomatology, and anxiety are more frequent in cannabis-induced psychosis. On the other hand, scores of negative psychotic symptoms have been found to be lower in cannabis-related psychosis compared to psychosis without cannabis use (3). In addition, patients with cannabis-induced psychosis are less neurodevelopmentally vulnerable and perform better in verbal learning, memory, social cognition, and processing speed compared to other psychotic patients (4). But in the long term, it has been shown that continued cannabis use has deteriorating effects on cognitive function. In a neuroimaging study, chronic cannabis use was associated with decreased amygdala and hippocampus volume (5).

Childhood trauma is a well-established risk factor for both drug abuse and psychosis. A recent review showed that psychotic patients exposed to childhood trauma have earlier onset, worse prognosis, and worse outcomes. Moreover, cannabis use increases the risk of psychosis by four times in people exposed to childhood trauma (6). One might say that childhood trauma creates vulnerability for psychosis, and later cannabis use can pull the trigger.

Highlights

• Negative psychotic symptoms were more apparent in participants without cannabis use.

• Cognitive functions were slightly better in cannabis user participants.

• There were no differences regarding positive psychotic symptoms between two groups.

Available data suggests that psychosis with a history of cannabis use has a different clinical presentation and prognosis. Psychotic patients who use cannabis have fewer impaired cognitive functions, but all of this data is limited (7). Most of the studies included multiple substance user patients (8–10). In addition, there are very limited studies with antipsychotic-naive patients. In this study, we aimed to investigate the suggested differences in psychotic symptoms, treatment responses, and cognitive functions in first-episode psychotic patients. We tried to assess cognitive functions prospectively after treating the psychotic episode. Moreover, we attempted to determine how childhood trauma is related to psychosis and cannabis use. We hypothesized that positive psychotic symptoms would be more severe and negative psychotic symptoms would be less severe in the cannabis user group compared to the non-user group. Also, we hypothesized that the cannabis user group would perform better in cognitive tests compared to the non-user group.

METHODS

Participants

This study was approved by the Ege University Clinical Research Ethics Board (Approval Date: 18/09/2019; Number: 19-9.1T/46) and complied with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, revised in 2008. All participants have given written informed consent before enrollment. All patients with psychotic symptoms who referred to Ege University Emergency Clinic or Ege University Psychiatry Outpatient Clinic between December 2019 and August 2020 and met the inclusion and exclusion criteria were included in the study consecutively after evaluation. The definition of psychotic symptoms was made according to the Structured Clinical Interview for Diagnostic Statistical Manual-5 (SCID-5).

Inclusion Criteria were: 1) age above 18 and below 40, 2) experiencing psychotic symptoms in the last month, and 3) approval to participate after being informed about the study. Exclusion criteria were: 1) regular substance use other than cannabis, 2) clinically determined mental retardation, 3) medical illness which can affect cognitive functions, 4) experiencing psychotic symptoms for more than one month, and 5) history of psychiatric illness with psychotic features. Participants were divided into two groups depending on cannabis use. Cannabis use was determined by the statements of the patients and relatives. Participants stating cannabis use at least once a month were included in the cannabis user group, and patients with no history of substance use were included in the non-user group.

Clinical Assessment

SCID-5 was employed to make the psychiatric diagnoses and identify the comorbidities of the participants. Case report forms were filled out for the sociodemographic characteristics, psychiatric history, and family history of the participants. After the first interviews, the Scale for the Assessment of Positive Symptoms (SAPS) and the Scale for the Assessment of Negative Symptoms (SANS) were applied to examine psychotic symptoms. SAPS has four subscales consisting of; “Hallucinations”, “Delusions”, “Bizarre Behaviour”, and “Positive Formal Thought Disorder”. SANS has five subscales consisting of; “Flattened Mood”, “Alogia”, “Social Isolation and Anhedonia”, “Loss of Energy and Will”, and “Attention”. In these scales, higher scores indicate higher psychotic symptoms. The Drug Use Disorders Identification Test (DUDIT) was applied to identify the characteristics of substance abuse. DUDIT is an 11-item scale ranging from 0 to 44, and 10 points can be used as a cut-off for problematic substance use. The SAPS and SANS were re-applied to inpatients during their discharge and outpatients after a month to reevaluate their psychotic symptoms. Childhood Trauma Questionnaire (CTQ) was applied in this second interview. All the scales were valid and reliable in the Turkish population (11–18).

Neuropsychological Assessment

Trail-making test (TMT) A and B, Stroop Color and Word Test (SCWT), and Rey Auditory-Verbal Learning Test (RAVLT) were applied to participants who were inpatients during their discharge and outpatients after a month to evaluate their cognitive functions. TMT A and B primarily assess visual scanning, processing speed, and executive functions. SCWT was used to evaluate selective attention and processing speed. RAVLT was used to measure short-term auditory-verbal memory, rate of learning, and retention of information. All the cognitive tests were valid and reliable in the Turkish population (19–24).

Statistical Analysis

Statistical analyses were performed using Statistical Package for Social Sciences (SPSS) for PC (version 25) software. The Shapiro-Wilk test was used to assess normal distribution. Numeric variables were expressed as mean and standard deviations if there was a normal distribution and as median with interquartile range (IQR) if there was skewed distribution. Parametric comparative analyses for demographic and clinical characteristics of the groups were performed with Independent T Test; for non-parametric distributions, a Mann-Whitney U-Test was performed. Categorical data were analyzed using Chi-Square or Fisher’s Exact Test where appropriate. To explore the correlation between numerical variables, the Pearson correlation coefficient (r) was calculated in parametric distributions, and the Spearman correlation coefficient was calculated in non-parametric distributions. ANCOVA was performed to control the effect of covariates over variables. Sociodemographic characteristics, DUDIT scale score, and CTQ score were used as independent variables. SAPS, SANS, SCWT, RAVLT, and TMT A and B scores were used as dependent variables. The level of significance was set at p≤0.05. To determine the sample size, we performed an a-priori power analysis. When the following values were used; effect size f=0.08, α=0.05, and power (1-β)=0.8, we measured that 20 participants for each group should be recruited with a total of 40 participants.

RESULTS

A total of 46 patients were recruited for the study. However, as eight participants (six of them were cannabis users) did not attend the second interview, we excluded their data. Therefore, the sample consisted of 38 patients, 18 in the cannabis user group and 20 in the non-user group. The sociodemographic and clinical characteristics of the participants are shown in Table 1. Cannabis usage characteristics are shown in Table 2. The mean DUDIT score was 21.78±9.48 (minimum: 8 and maximum: 37) for the cannabis user group.

Table 1.

Sociodemographic and clinical characteristics of the sample

Sociodemographic and clinical characteristics		Cannabis-user group (n=18)	Non-user group (n=20)	Significance
Age, Median (IQR)¹		23.5 (11.25)	24 (12)	p=0.803 u=171.50
Gender, n (%)²	Male	13 (72.2%)	11 (55%)	p=0.272 χ²=1.208
	Female	5 (28.8%)	9 (45%)
Marital status, n (%)³	Married	2 (11.2%)	7 (35%)	p=0.130
	Single	16 (88.8%)	13 (65%)
Educational status, n (%)³	Primary school	0	2 (10%)	p=0.464
	Elementary school	5 (27.7%)	3 (15%)
	High school drop-outs	3 (16.7%)	2 (10%)
	High school graduates	3 (16.7%)	7 (35%)
	College	7 (38.9%)	6 (30%)
Occupational status, n (%)³	Unemployed	7 (38.9%)	10 (50%)	p=0.691
	Student	3 (16.7%)	4 (20%)
	Employed	8 (44.4%)	6 (30%)
Psychiatric comorbidities, n (%)²	Yes	12 (66.6%)	7 (35%)	p=0.051 χ²=3.80
	No	6 (33.3%)	13 (65%)
Family history of psychiatric disorder, n (%)²	Yes	9 (50%)	13 (65%)	p=0.350 χ²=0.874
	No	9 (50%)	7 (35%)
Family history of substance abuse, n (%)³	Yes	5 (27.8%)	0 (0%)	p=0.017
	No	13 (72.2%)	20 (100%)
Living with parents, n (%)³	Yes	13 (72.2%)	20 (100%)	p=0.017
	No	5 (27.8%)	0 (0%)

¹Mann-Whitney U test; p values, median values and interquartile ranges are given.

²Chi-Square test, p values are given.

³Fisher’s exact test, p values are given.

IQR: Interquartile range; n: Number

Table 2.

Substance use data of the cannabis user group

Variables		n (%)
Frequency of use	Once a month	2 (11.1%)
	Once a week	4 (22.2%)
	More than once a week	5 (27.7%)
	Daily Use	7 (38.8%)
Quantity of use (in a month)	1–2	2 (11.1%)
	3–4	4 (22.2%)
	5–6	5 (27.7%)
	>7	7 (38.8%)
Adolescent use	Yes	14 (77%)
	No	4 (23%)
Family history of cannabis use	Yes	5 (27%)
	No	13 (73%)
Alcohol use	Yes	18 (100%)
	No	0 (0%)
Tobacco use	Yes	17 (94.4%)
	No	1 (5.6%)

In the cannabis user group, 14 (71.5%) participants were diagnosed with substance-induced psychotic disorder, and four (28.5%) participants were diagnosed with substance-induced bipolar disorder with a manic episode featuring psychotic symptoms. Regarding comorbidities, four (28.5%) of the participants had attention deficit hyperactivity disorder (ADHD), three (21.4%) had alcohol use disorder, and one (7.1%) participant had persistent motor tic disorder. In the non-user group, 15 (75%) participants were diagnosed with brief psychotic disorder, three (15%) were diagnosed with bipolar disorder with manic episode featuring psychotic symptoms, and two (10%) of the participants were diagnosed with depressive episode with psychotic symptoms. In the non-user group, four (20%) participants had obsessive-compulsive disorder (OCD) and three (15%) participants had autism spectrum disorder (ASD).

Psychotic Symptom Assessments

There were no statistically significant differences between the two groups regarding SAPS scores applied in the first and second interviews, apart from the Positive Formal Thought Disorder subscale in the second interview (p=0.034, u=110.50). Collected data for psychotic symptoms are summed up in Table 3.

Table 3.

Psychotic symptom assessments of the sample

	Cannabis-user group (n=18)	Non-user group (n=20)	Significance
First interview SAPS score
Hallucinations¹	13.72±8	12.10±8.77	p=0.557 t=-0.593
Delusions¹	21.50±8.12	20.30±5.53	p=0.595 t=-0.537
Bizarre Behavior¹	5.50±3.22	5.45±2.70	p=0.959 t=-0.052
Positive Formal Thought Disorder¹	14.28±7.56	13.40±9.00	p=0.748 t=-0.323
Total¹	54.44±19.15	51.10±21.1	p=0.610 t=-0.509
First interview SANS score
Flattened Mood²	4.00 (11.50)	13.00 (18.75)	p=0.032 u=108.00
Alogia²	0.00 (5.00)	5.00 (12.25)	p=0.030 u=108.00
Loss of Energy and Will²	6.00 (4.00)	13.50 (8.75)	p=0.002 u=74.50
Social Isolation and Anhedonia²	7.50 (9.50)	20.00 (10.00)	p=0.001 u=59.50
Attention¹	9.28±4.11	9.00±3.17	p=0.816 t=-0.234
Total¹	31.94±13.16	58.35±23.78	p=0.001 t=4.167
Second interview SAPS score
Hallucinations²	0.00 (8.00)	0.00 (5.00)	p=0.866 u=175.00
Delusions²	4.00 (7.00)	5.00 (10.00)	p=0.550 u=160.00
Bizarre Behaviour²	0.00 (1.00)	0.00 (2.00)	p=0.329 u=157.50
Positive Formal Thought Disorder²	5.00 (7.00)	0.00 (4.00)	p=0.034 u=110.50
Total²	10.50 (14.50)	6.00 (19.00)	p=0.418 u=152.50
Second interview SANS score
Flattened Mood²	5.00 (11.25)	6.00 (8.25)	p=0.62 t=0.499
Alogia²	0.00 (3.25)	2.00 (4.00)	p=0.176 u=136.00
Loss of Energy and Will¹	4.94±3.31	7.45±4.71	p=0.069 t=1.874
Social Isolation and Anhedonia¹	7.67±5.12	13.71±6.61	p=0.004 t=3.117
Attention²	1.50 (3.50)	3.00 (2.75)	p=0.095 u=126.00
Total¹	22.06±12.60	35.05±19.65	p=0.022 t=2.395

¹Independent groups t test; p values, mean values and standard deviations are given.

²Mann-Whitney U test; p values, median values and interquartile ranges are given.

SAPS: Scale for the Assessment of Positive Symptoms; SANS: Scale for the Assessment of Negative Symptoms.

When the differences in the SAPS and SANS scores between two interviews were examined, there was no significant difference between the cannabis user and the non-user group for SAPS (Δ=-42.44±21.31 vs. Δ=-39.85±2.98 respectively; p=0.714). However, in the non-user group, difference in SANS score between two interviews was significantly higher (Δ=-9.88±14.66 vs. Δ=-23.3±22.48 respectively; p=0.038).

There was a medium positive correlation between the DUDIT score and the first interview SANS score (r=0.530, p=0.06). Drug use disorders identification test score was also strongly correlated with the first interview Flattened Mood subscale (r=0.726, p=0.001). No other significant correlations were found between the DUDIT and SANS/SAPS scores.

Cognitive Functions

When we compared the cognitive functions of two groups, we did not determine a cut-off point for the SAPS and SANS. Cognitive tests were performed one month after the first interview in outpatients and during discharge in hospitalized patients. Descriptive statistics for SAPS and SANS scale during the time of the cognitive tests are stated as follows (SAPS: 11.81±10.72, minimum: 0, maximum: 38; SANS: 28.89±17.73, minimum: 0, maximum: 84). There was a significant difference for only Stroop Color Reading Time (p=0.007, t=2.851) among all the cognitive function tests (Table 4). This significance remained after controlling education covariance (p=0.018). Correlations between the DUDIT score and cognitive functions were analyzed, and a positive correlation was found between the DUDIT score and “TMT B-A” (r=0.580; p=0.012). A negative correlation was also found between the DUDIT score and the RAVLT correct recognition count (r=-0.560; p=0.016).

Table 4.

Cognitive function assessments of the sample

	Cannabis User Group (n=18)	Non-User Group (n=20)	Significance
SCWT word reading test, sn¹	32.21 (8.27)	35.82 (9.78)	p=0.090 u=122.00
SCWT color reading test, sn²	81.76±21.35	109.48±35.90	p=0.007 t=2.851
SCWT number of non-corrected errors¹	2.50 (4.25)	3.00 (6.00)	p=0.690 u=166.50
SCWT number of spontaneous correction¹	3.00 (2.50)	2.00 (3.75)	p=0.431 u=153.50
TMT A, sn	49.27±14.25	51.96±19.18	p=0.631 t=0.486
TMT B, sn¹	110.48 (48.32)	137.13 (82.05)	p=0.179 u=134.00
TMT B-A, sn¹	57.44 (44.73)	80.82 (54.28)	p=0.136 u=129.00

¹Mann-Whitney U test; p values, median values and interquartile ranges are given.

²Independent groups t test; p values, mean values and standard deviations are given. SCWT: Stroop Color Word Test; TMT: Trail Making Test.

Childhood Trauma

There were no significant differences regarding CTQ scores between cannabis user and non-user groups. The collected data are shown in Table 5.

Table 5.

Childhood trauma data of the sample

CTQ	Cannabis-User Group (n=18)	Non-User Group (n=20)	Significance
Emotional abuse¹	8.50 (5)	8.50 (8)	p=0.580 u=161.50
Physical abuse¹	5.50 (3)	5.00 (3)	p=0.899 u=176
Emotional neglect	13.44±4.65	13.35±4.23	p=0.948 t=– 0.650
Physical neglect²	9.00 (4)	8.00 (3)	p=0.345 u=148
Sexual abuse¹	6.50 (4)	5.00 (4)	p=0.521 u=159.50
Total¹	42.00 (18)	43.00 (22)	p=0.942 u=177.50

¹Mann-Whitney U test; p values, median values and interquartile ranges are given.

²Independent groups t test; p values, mean values and standard deviations are given. CTQ: Childhood Trauma Questionnaire.

DISCUSSION

Our main finding in this study is the difference in negative psychotic symptoms between cannabis user and non-user groups. In the non-user group, negative psychotic symptoms were significantly higher than in the cannabis user group. As negative symptoms are linked with neurodevelopmental impairment, this finding is consistent with the evidence indicating a neurodevelopmental difference between psychosis in cannabis users and non-users. Patients with a history of cannabis use could be less neurodevelopmentally vulnerable, and there is a possibility that if they did not use cannabis, they might not have become psychotic (25). Cannabis use is associated with risk-taking behavior and novelty-seeking (26). Moreover, acquiring substances mainly involves social interaction. There can be a premorbid difference between two groups, and this distinction may cause the difference in negative symptom scores at the beginning of the disorder. Manrique-Garcia et al. showed that psychotic patients who continued using cannabis in the long term had more negative symptoms than patients who did not (27). In the present study, although the final SANS score was higher, it improved more in the non-user group than cannabis user group with the help of the treatment. This can be interpreted as there is a difference in treatment response between two groups, and cannabis use can worsen the prognosis. There is substantial evidence associated with cannabis-related treatment resistance in psychosis (28). However, since all of the participants were first-episode psychotics, using antipsychotics for the first time, leading to affective bluntness, etc., could have an effect on negative symptoms in the cannabis user group.

It is known that the severity of cannabis use is associated with worse prognosis, symptom severity, and treatment resistance (29). In line with the literature, we found a correlation between the severity of cannabis use and negative symptoms prior to the treatment. However, this relationship lost its significance after the treatment. In the literature, cannabis use is linked with positive psychotic symptoms, auditory hallucinations, and disorganized behavior (30). In our study, total SAPS score and all subscale scores were higher in the cannabis user group, but the differences were not significant between two groups apart from the Positive Formal Thought Disorder subscale in the second interview. This discrepancy can be related to the small sample size.

The sociodemographic characteristics of our sample were similar to the literature. There were no differences regarding age, sex, and education between cannabis user and non-user groups. Although there is evidence about earlier age of onset for psychosis with cannabis use (31), we did not find a significant difference between two groups in this study. This discrepancy again can be due to the small sample size. In line with the literature, family history of cannabis use and living separately from parents were associated with cannabis use. There was an almost marginally significant difference (p=0.051) for psychiatric comorbidities between the two groups. In the cannabis user group, psychiatric comorbidities were more frequent. When we compared the participants who started using cannabis during adolescence and after adolescence, there was a significant difference in age of onset of psychosis. However, only four participants in our sample started cannabis use after adolescence. Therefore, our findings should be investigated in further studies with larger sample sizes.

Cannabis has a significant effect on cognitive functions, especially in adolescence, as the synaptic pruning process occurs at the same time. In a recent meta-analysis, cannabis user psychotic patients outperformed schizophrenic patients in cognitive tests, but the results were clinically insignificant (32). Another study found that cannabis-induced psychotic patients have better cognitive functions than non-cannabis-related psychotic patients, but they concluded that the results were due to better premorbid IQ and neurodevelopmental resilience (33). In our study, the cannabis user group performed significantly better in Stroop Word Reading Test. This can indicate that cannabis user psychotic patients are superior in sustaining attention and inhibiting disruptive stimuli. In most of the remaining cognitive tests, the cannabis user group performed better, but the results were not statistically significant. This may be due to the sample size, which was the most important limitation of our study. Similarly, in patients with higher DUDIT scores, cognitive performance was more impaired, but again the results did not reach significance.

Childhood trauma is a known risk factor for both psychosis and substance abuse, and some studies show that cannabis use in people who were exposed to childhood trauma creates significant vulnerability for psychosis (34). In our research, we could not find a similar correlation. This may be due to cultural differences, using a self-report scale, or the small sample size.

Our study has some limitations. Most importantly, we have a relatively small sample size, affecting the effect size. We excluded patients using multiple substances because other substances have different effects on cognition and psychotic symptoms. But it was difficult to find people who solely use cannabis. Moreover, strict COVID-19 restrictions at the beginning of the pandemic led to a major difficulty in recruiting patients for the study. Identifying cannabis use based on the statements of patients and relatives was another limitation. We had aimed to do a drug test before the recruitment stage, with the support of the Ege University Office of Scientific Research Project. However, COVID-19 restrictions prevented us. Not excluding patients with comorbid psychiatric diagnoses can be interpreted as a limitation regarding cognitive functions. Not performing IQ tests is another limitation, and it might have an effect on the cognitive tests, but education level was controlled in statistical analysis. Finally, some patients were hospitalized and treated as inpatients between the first and the second interview. However, the remaining were outpatients and treated by different psychiatrists, so the antipsychotic drugs and dosages used were different. Also, the timing of the second interview was different between hospitalized and outpatient participants This, potentially affected the SAPS and SANS scores applied in the second interview. While we did not determine a cut-off point for SAPS and SANS scores before performing the cognitive tests, according to Andreasen et al. (35), scores lower than three for selected items on scales can be used as a remission indicator. In our sample, all the participants met this criterion for SAPS.

To our knowledge, contrary to most of the studies, we performed a structured interview by a trained psychiatrist with all the patients. All the comorbidities were considered. All patients were between 18 and 40 years old, so age was not a confounding factor for cognitive functions. Excluding patients who use other substances provided a chance to create a more homogenous group. Moreover, cognitive tests were performed during the second interview; therefore, the psychotic symptoms were more settled, and measurements were done more accurately.

In conclusion, our results may indicate that these two phenomena could differ in many ways. Therefore, different prevention strategies and treatment options should be considered for psychosis in cannabis users and non-users.