Increase in schizophrenia incidence rates: findings in a Canadian cohort born 1975-1985

Background. Results from previous studies on the incidence rates for schizophrenia are inconsistent, with some showing a declining rate [e.g., Suvisaari et al. (1999) Arch Gen Psychiatry 56:733-740] and others showing an increasing rate [e.g., Boydell et al. (2003) Br J Psychiatry 182:45-49]. Objective. This study examines (1) whether incidence rates are changing, (2) relationships amongst changing incidence rates and age, period and cohort effects, and (3) the impact of rate changes on rate projections. Design. A care-based cohort study carried out in British Columbia, Canada, 1989-1998. Bayesian statistical analyses were used to estimate rates and describe secular effects. Classical tests of significance were used to assess the relative importance of age, period and cohort effects. Results. Between 1989 and 1998, median rates per 100,000 persons changed from 77.1 (90% credible interval (CI): 42.1-137.7) to 89.9 (90% CI: 80.1-100.1) in females, and from 66.6 (90% CI: 38.8-113.3) to 119.6 (90% CI: 107.4-132.4) in males. Age effects were active for both males and females. Period and cohort effects were stronger for males than females. Conclusion. In contrast to most previous studies, we found an increasing incidence of schizophrenia. Precise projections of schizophrenia incidence beyond 5 years require large sample sizes over prolonged periods of follow-up.