A discordant twin study of premorbid cognitive ability in schizophrenia.

Twin studies are advantageous because sources of genetic and environmental variation are equated in ways that are not possible in standard case-control designs. We examined premorbid cognitive ability by comparing Armed Forces Qualification Test scores administered at the time of military enlistment in 21 schizophrenia-discordant twin pairs and 860 matched control twins. Scores were significantly lower in schizophrenia probands than in their nonpsychotic co-twins; co-twins were midway between probands and control twins. Effects were reduced when the discordant pairs were extended to include 33 psychosis-discordant pairs. Compared with controls, education at enlistment was significantly lower in psychosis probands and in co-twins from both schizophrenia- and psychosis-disordant pairs. Co-twins from psychosis-discordant pairs had significantly lower education at midlife than controls. Results suggest that cognitive ability is influenced by familial vulnerability for schizophrenia or psychosis, and that premorbid cognitive ability is lower in schizophrenia versus psychosis in general. Educational advancement may be slightly slowed by this familial vulnerability, but results were equivocal with regard to attenuation of one's ultimate educational attainment.