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 disorders, and that being male, of younger age and current alcohol use were associated with ongoing cannabis use.

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.