Potential genetic variants in schizophrenia: a Bayesian analysis.

A number of different gene polymorphisms have been found to dispose for the development of schizophrenia. However, no single gene polymorphism is sufficient for the precipitation of schizophrenia. Swedish psychosis patients (n=103) and control subjects (n=89) were analyzed for 36 single nucleotide polymorphisms in 30 candidate genes for schizophrenia. Evidence of association was analyzed with Bayesian statistical methods. Variants in the genes coding for dopamine-D2 receptor, brain-derived neurotrophic factor (BDNF), neuropeptide Y (NPY), neuregulin 1, reelin and synapsin 3 showed association with schizophrenia, although few subjects were found in the minority allele for the two latter variants. The six gene variants, all with suspected connection to schizophrenia, were found to be risk factors when considered in combination, but not separately. The results indicate that the Bayesian statistical method gives additional possibilities in the search for risk factors for schizophrenia or other complex disorders.