The cognitive performance of 47 elderly psychotic patients with onset
of symptoms in late life (late paraphrenia) was compared to that of 33
controls matched for age, sex, ethnic origin, number of years of educ
ation, and pre-morbid IQ as measured by the NART. Neuropsychological i
ndices of general cognitive functioning (MMSE, CAMCOG, WAIS-R verbal a
nd performance scores) showed that patients were performing the tasks
at a significantly lower level than controls. Patients also showed a t
rend to have a lower span capacity than controls, particularly at the
spatial span subtest. There was no obvious impairment of learning as m
easured by the digit and spatial recurring span tasks nor of simultane
ous matching-to-sample ability. However, patients' performance on a de
layed-matching-to-sample procedure was significantly worse than that o
f controls. In addition, patients performed worse than controls on the
Recognition Memory Test for Faces, but not for Words. Finally, the pe
rformance of patients on tests assessing executive functioning (Verbal
Fluency Test, Computerized Extra and Intra-Dimensional Shift Task, Co
mputerized Spatial Working Memory Task, and Computerized Tower of Lond
on Task) was consistently worse than that of controls. These results s
uggest that psychotic states arising in late life are predominantly as
sociated with a decline on measures of general cognitive ability and e
xecutive functioning. The neuropsychological meaning of these findings
is discussed in the light of cognitive models of psychotic symptoms,
as well as of schizophrenia and dementia research. We concluded that t
he lack of a clear pattern of impairment among these patients may be t
he result of their clinical and cognitive diversity.