Letters to the Editor
Beyond Principal-Component Analysis of the Positive and Negative Syndrome Scale in Patients With Schizophrenia
Dear Editor:
Canciel and others (1) add their study of Positive and Negative Syndrome Scale (PANSS)-rated symptoms of schizophrenia to the growing literature indicating that 5 symptom factors are necessary to account for symptoms of the disorder. They note many similarities in the items contained in each factor across studies but also allude to several differences that may be related to the study sample’s clinical characteristics. Other possible reasons for differences in factor composition may be related to sample size and to the limitations of the principal-component method. Indeed, principal-component factor analysis is recognized as an exploratory method. We wish to bring attention to our large-scale, multicentre study of the PANSS factor structure. In it, we use the more rigorous method of confirmatory factor analysis (2). In a sample of 1233 subjects with schizophrenia, we found that groups varying widely in age, chronicity, and illness phase did not significantly differ in their symptom structure. We identified a 5-factor structural model of the PANSS that met statistical criteria for good model fit. The criterion of good fit is an index of the degree of correspondence between the order in sample data and in the proposed model. Our model used 25 of the 30 PANSS items, organized into 5 factors: negative, positive, activation, dysphoric mood, and autistic preoccupation. Although identification of a good-fit model establishes the model’s internal consistency, validity studies are necessary to demonstrate model utility. The model’s discriminant validity has been demonstrated in a study of sex differences in the relation of homelessness to symptom severity (3), in a study of symptom differences between familial and deficit-syndrome schizophrenia (4), and in a study of the association between symptoms and cognitive-perceptual deficits in schizophrenia (5). Symptom subscales based upon this model have been published, as have tables of norms based upon an adult population with chronic schizophrenia (6).
References
1. Canceil O, Sampaio-Meireles M, Poirier-Littre MF, Bourdel MC, Olie JP, Attar-Levy D, and others. Principal-component analysis of the Positive and Negative Syndrome Scale in patients with schizophrenia: does a 5-factor model apply to published data? Can J Psychiatry 2002;47:97.
2. White L, Harvey PD, Opler L, Lindenmayer JP. Empirical assessment of the factorial structure of clinical symptoms in schizophrenia. A multisite, multimodel evaluation of the factorial structure of the Positive and Negative Syndrome Scale. The PANSS Study Group. Psychopathology 1997;30:263–74.
3. Opler LA, White L, Caton CL, Dominguez B, Hirshfield S, Shrout PE. Gender differences in the relationship of homelessness to symptom severity, substance abuse, and neuroleptic noncompliance in schizophrenia. J Nerv Ment Dis 2001;189:449–56.
4. Malaspina D, Goetz RR, Yale S, Berman A, Friedman JH, Tremeau F, and others. Relation of familial schizophrenia to negative symptoms but not to the deficit syndrome. Am J Psychiatry 2000;157:994–1003.
5. Doniger GM, Silipo G, Rabinowicz EF, Snodgrass JG, Javitt DC. Impaired sensory processing as a basis for object-recognition deficits in schizophrenia. Am J Psychiatry 2001;158:1818–26.
6. Kay SR, Opler LA, Fiszbein A. Positive and Negative Syndrome Scale (PANSS). Toronto: Multi Health Systems; 2000.
Leonard White, PhD
West Brentwood, New York
Lewis A Opler, MD, PhD
Mount Vernon, New York
|
