Cumulative effect of anatomical risk factors for schizophrenia: An MRI study

Background: Although schizophrenic and control subjects differ on a variety of neuroanatomical measures, the specificity and sensitivity of any one me asure for differentiating between groups are low. This study investigated t he cumulative effect of deviant brain structure on diagnosis. Methods: Hemisphere and third ventricle volume and the normalized (Talairac h) location of three association, cortex sulcal landmarks were measured on high-resolution MRI scans in 37 male patients with schizophrenia and 33 mal e control subjects matched on age, handedness, and parental socioeconomic s tatus. Results: While there were few group differences on individual anatomical me asures, the 10 variables reliably discriminated between the two groups when used in concert in a discriminant function analysis (F[10,59] = 3.6, p < . 0009) with 77% of the subjects correctly classified. Five of the measures ( left posterior cingulate, left inferior frontal sulcus, right sylvian fissu re, and left and right halves of the third ventricle) correlated significan tly with the discriminant function (p < .005). Conclusions: This is the first study to demonstrate that schizophrenics can be distinguished from matched controls on the basis of brain anatomy alone . The risk of schizophrenia may depend on the total amount of neural devian ce, rather than on anomalies in a single structure or circuit. Biol Psychia try 1999;46:374-382 (C) 1999 Society of Biological Psychiatry.