W. Loffler et H. Hafner, Replication of the 3-factor model by Liddle to a population-related first-episode sample of schizophrenic patients, NERVENARZT, 70(5), 1999, pp. 416
The issue of this study was the investigation of the dimensional structure
of non-psychotic and psychotic symptoms in 232 first-episode schizophrenic
patients (ICD-9 295., 297., 298.3, 298.4). The study was conducted within t
he ABC-Schizophrenia-Study. The three-factor-model of Liddle with three fac
tors (psychomotor poverty, disorganisation, reality distortion) was replica
ted for the time at first admission. The model is also valid for first-epis
ode-patients as well as to chronic patients. The comparison of the three-fa
ctor-model of Liddle with Crow's dual process model, Andreasen's bipolar mo
del and the "severity-liability" model was done by means of confirmatory fa
ctor analysis. The comparison shows that at first admission, the three-fact
or-model fitted in best with the data. In contrast to previous analyses wit
hin the ABC-Study, in which positive correlations have been found between p
ositive and negative symptoms, no positive correlation exists between Liddl
e's negative and positive dimensions. This may be the consequence of the su
bdivision of the positive dimension into the two dimensions disorganisation
and psychotic symptoms. As within the three-factor-model only the negative
dimension and disorganisation correlated weekly, the three dimensions are
best viewed as relatively independent for the time at first admission. Ther
e are no associations between sex, type of onset, age at onset and the thre
e dimensions of Liddle's model. Patients with the familial load are more di
sorganized and patients with obstetric complications show more negative sym
ptoms. While the negative dimension shows a high stability over five years,
the dimensions "disorganisation" and "positive symptoms" are not stable ov
er time. However, there is a high degree of correlation for the dimensions
"disorganization" and "positive symptoms" among cross-sections while the ne
gative dimension was independent of the other two dimensions. The negative
dimension is a highly significant predictor for social disability and socia
l development over five years, whereas the dimensions "disorganization" and
"positive symptoms" have no prognostic importance for the outcome in the l
ong term.