Related Experiment Videos
A multivariate prediction model of schizophrenia.
John W Carter1, Fini Schulsinger, Josef Parnas
1Social Science Research Institute, University of Southern California, Los Angeles 90089-0375, USA. jwcarter@almaak.usc.edu
Schizophrenia Bulletin
|June 11, 2003
Summary
Multivariate prediction models accurately identified schizophrenia risk using premorbid factors like genetic risk and disruptive school behavior. Early identification and prevention efforts can be enhanced by understanding these complex gene-environment interactions.
Area of Science:
- Psychiatry
- Developmental Psychology
- Behavioral Genetics
Background:
- Univariate models for schizophrenia prediction are limited in efficacy.
- Multivariate approaches are needed to improve predictive accuracy and identify key risk factors.
- Premorbid assessment offers a window into early developmental trajectories.
Purpose of the Study:
- To maximize the prediction of schizophrenia using multivariate analysis of premorbid measures.
- To evaluate the relative importance of various predictors for schizophrenia.
- To explore gene x environment interactions in schizophrenia development.
Main Methods:
- A longitudinal study of Danish subjects with high (n=212) and low (n=99) familial risk for schizophrenia spectrum disorders.
- Assessment of 25 premorbid variables across seven domains at age 15.
- Discriminant function analyses to predict schizophrenia diagnoses made 25 years later.
Main Results:
- Schizophrenia was predicted by the interaction of genetic risk with rearing environment and disruptive school behavior.
- These premorbid measures correctly predicted two-thirds of schizophrenia outcomes within the high-risk group.
- Prediction accuracy improved with higher genetic loading (two affected parents), supporting a multiplicative gene x environment model.
Conclusions:
- Multivariate premorbid assessment effectively predicts schizophrenia, highlighting the importance of gene-environment interactions.
- Disruptive school behavior and adverse rearing environments interacting with genetic risk are key predictors.
- Findings support the development of targeted early identification and primary prevention strategies for schizophrenia.