Clinical correlates of first episode early onset psychosis in KwaZulu-Natal, South Africa

  • Saeeda Paruk
  • Khadija Jhazbhay
  • Keshika Singh
  • Benn Sartorius
  • Jonathan K. Burns


Background: The study of first episode early onset psychosis can yield many clues to understanding the early development of psychosis and guide interventions to decrease psychosis risk and improve outcome. The aim of the study was to investigate the socio-demographic profile and clinical correlates in early onset psychosis.

Method: Forty-five adolescents with first episode early onset psychosis were assessed by a clinical interview, socio-demographic questionnaire, the Positive and Negative Syndrome Scale (PANSS), Symptom Onset in Schizophrenia (SOS) score, the World Health Organization Alcohol, Smoking and Substance Involvement Screening Test (WHO ASSIST) questionnaire for substance misuse and a urine cannabis level.

Results: The mean age of the respondents was 15.9 years (SD 1.8, range 10–18 years). Thirty-one (69%) were male. There was a poor recognition of the prodromal period in 22 (49%) patients and caregivers. The mean duration of untreated psychosis (DUP) was 27.2 weeks (SD 56.7). There was a negative correlation between DUP and age of onset (p < 0.05). Mean age at onset of psychosis (males 15.7 years, SD 2.2, and females 15.3 years, SD 2.6) and mean age of presentation (males 16.0 years, SD 1.8, and females 15.5 years, SD 1.7) was slightly younger in the females than males, but not statistically significant. A total of 25 (56%) adolescents reported lifetime cannabis use and 24 (96%) were male.

Conclusion: Symptoms of the prodromal period were poorly recognised. The mean DUP suggests a significant delay in treatment and younger children had a longer DUP. The gender differences in presentation (marginal) and substance use (highly significant) suggests that there may be different environmental risk factors for males and females in early onset psychosis or that early onset psychosis is more genetically homogenous and less dependent on environmental triggers.


Journal Identifiers

eISSN: 1728-0591
print ISSN: 1728-0583