Background. High expressed emotion (EE) in families reliably predicts relapse in schizophrenia leading to the assumption that high EE stresses patients and is the cause of relapse. Attempts at validating the stressful properties of high EE using autonomic skin conductance measures have been inconclusive. Since memory loaded vigilance tasks are sensitive to stress, we used a memory loaded version of the Continuous Performance Test (CPT) to see if test performance would change in the presence and absence of high or low EE relatives. If high EE was stressful, performance should decline in their presence of high EE relatives. If low EE was protective against stress performance should improve in the presence of low EE relatives.

Methods. The CPT was administered to 41 schizophrenic and 51 normal control participants. CPT testing was conducted in the absence, then presence, then absence of high or low EE relatives.

Results. Control participants had significantly greater CPT scores than schizophrenic participants; there was no effect of the relative's presence or absence on CPT performance for either the high or low EE group. Schizophrenic participants from high EE homes demonstrated significantly better, rather than worse, CPT scores than those in low EE homes.

Conclusions. These findings fail to support the notion of high EE as stressor, but show that high EE and low EE patient groups perform a vigilance task significantly differently. We propose that patients from low EE groups may be underaroused relative to their high EE counterparts and that this underarousal leads to less than optimal performance on the CPT.