The extent and causes of covariance between schizotypy and neurocognition is not well-known yet. Certain models conceive their association as necessary for the construct validity of schizotypy, whereas others view them as independently contributing to a multivariate endophenotype. It is also not clear whether those at increased genetic risk for schizophrenia present stronger covariance, reflecting an extra latent source of variance. We analysed their association within relatives of schizophrenia patients defined with FIGS as Presumed Carriers -PC- of the genetic risk for schizophrenia, Presumed Non Carriers -PNC-, and controls.

108 healthy relatives of schizophrenia patients and 72 healthy controls were assessed with the SCID-II and completed the SPQ-B. Neurocognitive assessment: Letter-Number Sequencing (LNS), WCST, CPT-IP, verbal fluency, and logical memory.

Partial correlations adjusting for age and education showed that within PC-relatives self-rated negative schizotypy was associated with lower LNS and CPT-IP; positive schizotypy was associated with CPT-IP, and disorganization with memory and failure to maintain set. Schizoid symptoms had an association with failure to maintain set (though not perseveration) and paranoid symptoms with memory. Within PNC-relatives, negative schizotypy was associated with lower verbal fluency and more perseverative errors. Within controls, positive schizotypy was associated with perseverative errors and both positive and negative dimensions were associated with verbal fluency.

Results indicate a wider array of covariation between relatives with presumed higher genetic liability. A consistent pattern of associations between psychotic-like dimensions and the brain functions tapped by neurocognitive tests did not emerge across groups.