How Age and Gender Predict Illness Course in a First-Episode Nonaffective Psychosis Cohort

Objective: Male gender and young age at onset of schizophrenia are traditionally associated with poor treatment outcome and often used to determine prognosis. However, many studies use nonincident samples and fail to adjust for symptom severity at onset. We hypothesized that age and gender would influence severity of presentation but would not predict outcome after adjustment for symptoms at presentation.

Method: 628 people with first-episode ICD-9 and DSM-IV nonaffective psychosis from 2 historical cohorts recruited from sequential presentations in Canada and the United Kingdom (1996-1998) were assessed prospectively at presentation and over 12-18 months using the Positive and Negative Symptom Scale (PANSS).

Results: Models of the age-at-onset distributions with 2 underlying modes at similar ages in women (ages 23 years and 47 years) and men (ages 22 years and 46 years) had relatively good fits compared to single-mode models (χ21 better by 9.2 for females, 8.0 for males, both P < .05). At presentation, scores for negative symptoms were 1.84 points worse for males (95% CI, 1.05 to 2.58; P < .001) in a mixed effects model. Younger age also predicted higher negative scores at presentation (partial correlation r = -0.18, P < .01; P < .001 in the mixed effects model). Findings were similar for cognitive-disorganized symptoms. However, after controlling for baseline symptoms, age at onset and gender did not significantly predict subsequent symptom course in the mixed effects models.

Conclusions: Gender and age at onset are independently associated with symptoms at presentation but not with medium-term course of schizophrenia. This finding reinforces the importance of early identification and prevention of severe negative symptoms at first episode, whatever an individual’s age and gender.