Meta-analytic approaches to determine gender differences in the age-incidence characteristics of schizophrenia and related psychoses

Dan Jackson, James Kirkbride, Tim Croudace, Craig Morgan, Jane Boydell, Antonia Errazuriz, Robin M Murray, Peter B Jones

Research output: Contribution to journalArticlepeer-review


A recent systematic review and meta-analysis of the incidence and prevalence of schizophrenia and other psychoses in England investigated the variation in the rates of psychotic disorders. However, some of the questions of interest, and the data collected to answer these, could not be adequately addressed using established meta-analysis techniques. We developed a novel statistical method, which makes combined use of fractional polynomials and meta-regression. This was used to quantify the evidence of gender differences and a secondary peak onset in women, where the outcome of interest is the incidence of schizophrenia. Statistically significant and epidemiologically important effects were obtained using our methods. Our analysis is based on data from four studies that provide 50 incidence rates, stratified by age and gender. We describe several variations of our method, in particular those that might be used where more data is available, and provide guidance for assessing the model fit. Copyright © 2013 John Wiley & Sons, Ltd.
Original languageEnglish
Pages (from-to)36-45
JournalInternational Journal of Methods in Psychiatric Research
Issue number1
Publication statusPublished - 4 Mar 2013

Bibliographical note

Copyright © 2013 John Wiley & Sons, Ltd.

Cite this