Prevalence of cannabis use in people with psychosis in KwaZulu-Natal, South Africa

Background There is a high prevalence of cannabis use in patients with schizophrenia spectrum and other psychotic disorders, with comorbid cannabis use in this population being associated with poorer long-term outcomes. Aim To determine the prevalence of cannabis use in patients with a schizophrenia spectrum and other psychotic disorders. Setting The study was conducted at a psychiatric hospital in Durban, KwaZulu-Natal Province, South Africa. Methods A review of clinical records of patients admitted to the hospital for the period, June 2018 to June 2020, was conducted. Results A total of 370 clinical records were reviewed, of which 48.9% reported current and 51.1% lifetime cannabis use. Being male was significantly associated with current and lifetime cannabis use (OR = 4.90, 95% CI 2.49–9.62 and OR = 6.27, 95% CI 3.28–11.95, respectively). Current alcohol use was also associated with current cannabis use (CCU) (OR = 3.06, 95% CI 1.78–5.28), and age 45 years and older was associated with a lower odds of cannabis use (OR = 0.30, 95% CI 0.09–0.96). Forty-eight per cent of participants were admitted three or more times, and readmission was associated with cannabis use (p = 0.01). There was a lack of association between cannabis use, readmission and human immunodeficiency virus (HIV) status, after controlling for variables such as alcohol use and gender. Conclusion Almost 50% of people admitted with schizophrenia spectrum and other psychotic disorders have comorbid current and lifetime cannabis use. There is a need for dual diagnosis units to address comorbid substance use in people with psychotic disorders, as it leads to poorer outcomes. Contribution The study found that there is a high prevalence of cannabis use in people with psychosis. Therefore, it is imperative that we revise treatment programs in our psychiatric units and there is an urgent need for dual diagnosis programs that address substance use in this group of patients.


Introduction
Illicit substance use is a worldwide problem and contributes significantly to the global burden of disease. 1 The United Nations office on Drugs and Crime world drug report (2020) found that cannabis is one of the most commonly used drugs, with approximately 192 million users, of whom 35 million meet the criteria for cannabis use disorder. 2 Schizophrenia is a chronic and severe mental illness that is characterised by positive, negative, affective and cognitive symptoms. It runs an often-disabling course and places a considerable burden on patients, their families, the healthcare system and society at large. 3,4 Comorbid cannabis use in schizophrenia spectrum and other psychotic disorders is associated with a more severe course of illness. Cannabis users with psychosis have an earlier onset of illness, increased symptom severity, poorer adherence to psychiatric medication, poorer functional outcomes, higher relapse rates and longer hospital stays compared with non-users. 5,6,7,8,9,10,11 In addition, people with schizophrenia spectrum and other psychotic disorders and comorbid cannabis use tend to have poorer psychosocial functioning, homelessness, violence, incarceration and increased burden of care for families. 4,12,13,14 having been reported. 10 In a meta-analysis of 35 studies of adults presenting with first episode psychosis (FEP), 33.7% reported cannabis use, 30.7% of outpatients with FEP reported past month cannabis use and 34.7% met the criteria for lifetime cannabis abuse or dependence. 17 Studies conducted in the United States of America (USA) and Canada found that cannabis use is twice as common in patients with severe mental illness, such as schizophrenia, compared to the general population. 18 The prevalence of cannabis use in South African populations with psychotic illness has seen reported to be between 37% and 68%, depending on the setting and population. 19,20,21 A Cape Town study found the most common substance use disorder associated with psychosis was cannabis use disorder, with a prevalence of 34.3%. 15 Recent studies show that the use of high potency cannabis increases the risk of FEP and subsequent schizophrenia, and that psychosis and prodromal schizophrenia symptoms may also exacerbate cannabis use. 22 Genetics and environmental factors play a role in this bidirectional relationship between psychosis and cannabis use. 23 Various socio-demographic and clinical factors may be associated with cannabis use among people living with psychosis. International studies found that cannabis use in patients with psychosis was predominantly in males, younger in age, of a low education level and socio-economic status. 24,25 Data from South Africa (SA) and other African countries have shown similar findings, with unemployment, being single, an inpatient status, and recurrent admissions being additional factors associated with cannabis use. 5,15,16,19,26,27 Hence we note a high comorbidity between cannabis use and psychosis, and that the dual diagnosis of a cannabis use and psychotic disorder is a poor prognostic factor. A study conducted in India had similar findings where cannabis use had a high prevalence for psychotic disorder and this co-occurrence is a poor prognostic factor. 28 In KwaZulu-Natal (KZN) Province, SA, Paruk et al. 20 found a cannabis use prevalence rate of 68.8% in the adolescent psychotic population, while Burns and colleagues found a prevalence rate of 35% in the adult population of the same province. 29 These studies were all conducted about a decade ago, before the personal use of cannabis was decriminalised in SA. 30 The factors associated with comorbidity in the African setting may also differ from elsewhere, as this may be influenced by cultural and environmental differences between countries and availability of cannabis, while the decriminalisation and legalisation policies on cannabis may have an impact on the patterns of use.
This research aimed to describe the socio-demographic and clinical profiles of patients admitted with schizophrenia spectrum and other psychotic disorders to a psychiatric hospital. The study also aimed to determine the prevalence and pattern of cannabis use (current vs. lifetime) in this population and to describe any associations between cannabis use and socio-demographic and clinical variables.

Aim
The association between cannabis use and psychosis has been established in recent years, but there has been a move to decriminalise personal cannabis use globally and in SA; hence the authors aimed to review the prevalence of current and lifetime cannabis use in people living with psychosis.

Methods
A retrospective chart review was conducted at a tertiary psychiatric hospital that offers psychiatric services in the eThekwini District, KZN Province, SA.

Study population
All in-patients 13 years and older meeting the DSM-5 criteria for schizophrenia spectrum and other psychotic disorders who were admitted between 01 June 2018 and 31 June 2020 were included. Clinical records were obtained from the hospital registry of admissions, with those with incomplete or missing data being excluded from the study.

Data collection
A structured socio-demographic and clinical data questionnaire was used to collate data from the clinical records, the former including age, gender, marital status, education level, current employment status and place of residence. The clinical data collected included family history of psychotic and non-psychotic illness and substance use, Diagnostic and Statistical Manual of Mental Disorders (DSM-5) diagnostic categories, 31 age at first admission and duration of illness. Substance use history, current and lifetime use, duration of hospitalisation, number of hospitalisations and human immunodeficiency virus (HIV) status data were also collated. Cannabis use history was based on patient and family report, and the history of other illicit substances was also extracted. All data extracted from the clinical files was collected by the principal investigator, captured manually into a data collection sheet and later captured electronically onto Research Electronic Data Capture (REDCAP) database. 32 For the study, the following terminology is defined based on a review of the literature and consensus by the research team: • Schizophrenia spectrum and other psychotic disorders include schizophrenia, other psychotic disorders and schizotypal personality disorder and are defined by the presence of at least one of the following domains: hallucinations, delusions, disorganised thinking or behaviour and negative symptoms. 31 • Lifetime cannabis use: A positive report of lifetime cannabis exposure before or after seeking treatment.

Sample size
Davis et al. 34 in their study of patients with serious psychiatric illness such as schizophrenia in KZN, found the prevalence of cannabis use to be at 49.4%. Using the formula n = z 2 p(1−p)/d 2 (n: sample size, z: standard deviation set at 95% standard value of 1.96, p: prevalence and d: error margin), the researchers aimed to reach a target sample size of 349 participants, which was adequate to achieve a 95% confidence interval at an error margin of 5.25% in this study. 35

Ethical considerations
Ethical approval was obtained from the Biomedical Research Ethics Committee of the University of Kwa-Zulu Natal (BREC/00002640/2021) and the Kwa-Zulu Natal Department of Health. The study is a retrospective chart review and therefore did not require consent from participants.

Patient socio-demographic profile
A total of 370 patient clinical records were included in the study, with two being excluded because of inadequate information. The socio-demographic and clinical information is presented in Table 1; 66.2% of participants were between the ages 16 and 34 years; most were male (70.5%), unmarried (93.9%), had a highest level of education of grade school (88.0%) and were unemployed (89.2%). The median duration of the psychotic illness was three years; 48% had been admitted three or more times, while the current and lifetime cannabis use was at 48.9% and 51.1%, respectively.  with current and lifetime cannabis use. A history of three or more admissions was associated with lower prevalence CCU compared to no previous admission.

Regression results
The logistic regression models are presented in Table 3

Discussion
The main findings of this study were the high prevalence of current (48.9%) and lifetime (51.1%) cannabis use among people with schizophrenia spectrum and other psychotic Our study showed that CCU in psychosis is high at 48.9% but slightly lower compared with a study conducted in the Eastern Cape Province of SA, which showed a prevalence of 59.8%. 19 This difference may have been attributed to the Eastern Cape cohort being individuals with FEP only. Temmingh and Mall, in their Western Cape cohort, reported a lower prevalence of 34.3%, which may be because of their assessing for substance use disorder rather than cannabis use. 15 In the Western Cape Province (SA), 57% of adolescents presenting with FEP were reported to be cannabis users. 21 Paruk et al. 20 reported the prevalence of cannabis use at 68.8% in adolescents with FEP in KZN, while Burns et al. 36 reported 35% prevalence of comorbid cannabis use in their adult FEP sample. This study was also conducted in KZN but included patients 16-45 years of age, and it is possible that patients may be less likely to continue cannabis use with advancing age, as was evident in our study, where those 45 years and older had lower odds of cannabis use compared with the 16-34 years age group. This could be related to peer pressure and stress or negative life events experienced by younger people with psychosis. 37 The literature also suggests that cannabis exposure during the adolescence period, a critical time in neurodevelopment, may have lasting adverse effects that predispose to the development of psychosis as the individual grows older, even when they are no longer using cannabis 38,39,40 ; hence the finding of increased use in younger people is a concern and we need to ensure that young people at risk of or presenting with psychosis receive psychoeducation and care for comorbid substance use.     Similar to our findings, other studies have also consistently reported cannabis users in populations with psychosis to be predominantly male, of young age, having a low level of education and unemployed. 5,15,16,19,24,25,26 In a USA community hospital cohort, the findings were that being male, young, unemployed and homeless was significantly associated with cannabis use in patients with schizophrenia spectrum disorders. 41 A study conducted in Spain in patients with FEP also found users to be predominantly male who started cannabis use earlier in life. 42 A study conducted in Morocco also reported patients diagnosed with schizophrenia who were using cannabis were young, male, presented earlier in life and had poor adherence to medication. 43 Local findings have also been consistent, with cannabis users being more likely to be young and male. 5,19 The vulnerability of young males suggests the need for earlier public health programmes on substance use among this cohort who may be at risk of psychosis to improve outcomes.
Our study found that 52.8% of current cannabis users were also using alcohol, which is consistent with an international study conducted in Canada, where people with psychosis were likely to have an alcohol and cannabis use disorders at 21% and 27%, respectively. Rates of substance use were highest in the first five years of psychotic illness at 20% -33% for alcohol and 21% -45% for cannabis. Alcohol use was associated with higher social functioning and less negative symptoms, 10 this being similar in a study conducted in the Eastern Cape, SA; which found that patients with FEP used cannabis followed by alcohol use. 19 This is also supported by a study conducted in the USA, which found that patients with schizophrenia spectrum disorders and cannabis use were more likely to use other substances, such as alcohol, tobacco and cocaine. 41 It suggests the need to screen those with psychosis for polysubstance use and reinforces the call for the much-needed dual diagnosis units that cater to substance use rehabilitation and psychosis, which are not available in KZN, SA.
The finding that almost 50% of our study population had more than three or more admissions was also concerning. While on logistic regression this study did not find any associations between cannabis use and re-hospitalisation, suggesting that the findings were compounded by other factors such as poorer access to healthcare in the province where patients were possibly not admitted to hospital and managed in the community. This finding is, however, contrary to findings in the literature where studies have consistently found that schizophrenia and continued cannabis use leads to poorer outcomes, such as a more complicated course of illness, as evidenced by frequent relapses, shorter periods of time between relapses, longer hospitalisations and more intensive psychiatric care. 5,11,19,27 This is also consistent with both local as well as other international literature. 27,44,45 As the literature shows that discontinuing cannabis use has improved outcomes, it becomes essential that we review the modifiable factors associated with psychosis outcome. 11 There was a lack of association between the clinical variable HIV infection with cannabis use in people with psychosis in this study, which is consistent with another local study, but inconsistent with international research. 46,47,48 This may be because of almost 30% of participants not having a documented HIV result, which may be attributed to poor record keeping, limited HIV testing access or the inability to provide consent, which may have affected the finding. Substance use and psychosis may increase risky sexual behaviour and hence HIV risk 49,50,51 ; therefore, the association between HIV and cannabis use in those living with psychosis needs to be explored further.

Limitations
The limitations of this study include its retrospective nature, where the data rely on the quality of record keeping, which may not be consistent or detailed well. This was a hospital-based study, which may introduce sample bias and limit generalisability to community samples. In addition, cannabis use reporting was subjective from the patient and collateral information available and not verified by urine toxicology.
However, the strength of study is that this study reflects the persistent high prevalence of cannabis use and that young males remain most vulnerable. There is a need for programmes promoting greater awareness on potential harms of cannabis use in young people particularly those who may be at risk for psychosis.

Conclusion
The high rates of cannabis use, with almost half the people in the study with psychosis still reporting current use, is an important public health concern. Cannabis use was associated with a vulnerable group of young males with lower education and comorbid alcohol use, and with over 50% having more than three or more admissions; the data suggest that its continuous use leads to a poorer prognosis. The findings highlight the need to review treatment programmes offered at psychiatry units, with an urgent need for dual diagnosis programme that addresses substance use in people living with psychosis in a comprehensive and integrated manner.

Recommendations
Further longitudinal studies are required to better understand association between ongoing cannabis use and psychosis outcomes, including intervention studies. Clinically we need to address cannabis use with a more concerted effort and to develop dual diagnosis units to address comorbid substance use.