Püschel J, Stassen H H, Bomben G, Scharfetter C, Hell D
Psychiatric University Hospital Zurich, Research Department, Switzerland.
J Psychiatr Res. 1998 Mar-Apr;32(2):89-97. doi: 10.1016/S0022-3956(98)00046-6.
Based on a sample of 45 hospitalized, acute-schizophrenic patients and 45 carefully matched controls, we investigated the non-verbal characteristics of schizophrenic speech by means of an 'acoustic' speech analysis and determined the extent to which speaking behavior and speech sound characteristics had adjusted toward normal values at the time of hospital release. Using a multivariate discriminant function derived from a previous study of chronic schizophrenics, totally 77 (85.6%) individuals of our patient and control sample could be correctly classified by a set of 12 acoustic variables at entry into study. At hospital release, the majority of patients (64.4%) still exhibited speech impairment although acute psychopathology had significantly improved. A configuration of 6 acoustic variables, assessed at the time point of entry into study, predicted at high reliability the severity of the negative syndrome at hospital release. Acute medication effects did not explain these findings, thus underlining the potential diagnostic relevance of the speech analysis method. With respect to the relationship between speech characteristics and acute psychopathology throughout the time course of recovery, our results suggest that changes in speaking behavior and speech sound characteristics may be distinct aspects of schizophrenia that can persist in a subgroup of patients over a long period, mostly beyond the time point of hospital release. Accordingly, the speech analysis method might become very useful in detailing the nature and severity of deficits in patients after remission of positive symptoms.
