Patterns of psychophysiological and acoustic responses and flat affect in schizophrenia

Abstract

Flat affect in schizophrenia is usually regarded as a generalized decrease in emotion-related behaviors. However, some studies have suggested that this generalized concept has to be qualified by type of behavior and situation in which the behavior is embedded. We were also interested in investigating the relationship of overt emotion-related behaviors such as facial expression and voice variables with emotion specific autonomic response patterns in affective flattening. A study was done comparing 12 schizophrenics with flat affect, 12 schizophrenics with non-flat affect (DSM-III-R criteria) and a comparison group of 12 depressed patients on five psychophysiological measures and eleven voice measures under emotional manipulations of happiness, sadness and anger. Voice recordings were taken from a standard interview and three emotion monologues. Psychophysiological recordings were taken while the subjects were reliving these plus a neutral situation. The voice tapes were analyzed with Alpert’s VOXCOM computer program. Results show that duration of hospitalization was positively correlated with ratings of affective flattening, and so this variable was used as a covariate in all statistical tests. Chlorpromazine equivalent dose was not correlated to affective flattening. The emotion manipulation was tested with a MANOVA procedure for repeated measures. The following results pertain to the flat versus non-flat schizophrenic contrast: The flat affect schizophrenics showed significantly greater baseline heart rate compared to the non-flat schizophrenics under the anger condition. The following patterns were found for the contrast of the schizophrenics versus the depressed group: Flat affect schizophrenics showed greater corrugator EMG (eyebrow) activity than the depressed group across the emotions for one of two interactions. Also, pulse volume and the three autonomic measures together (pulse volume, heart rate, and respiration) show a significant group by emotion interaction effect for this contrast. In terms of acoustic variables, the groups showed significant baseline differences in mean gap length between utterances. The flat affect schizophrenics showed more variability in gap length, and they talked significantly less than the non-flat schizophrenics. With regard to the flat versus non-flat schizophrenic contrast, significant differences in group by emotion patterns were found in voice level, variability of gaps, and variability of voice level (for the sad condition only). For the schizophrenics versus depressed contrast, variability of voice level and percent talk time was significantly different across emotions. The group by emotion patterns for speech rate also differentiated the schizophrenics from the depressed group. Our results show that some emotion-related behaviors such as facial expression and some voice variables may vary in flat and non-flat affect schizophrenics depending on emotion. This suggests the need to assess affect in a variety of situations. This has implications for the rehabilitation and treatment of the individual schizophrenic. Discussion will focus on flat affect as a special case of the break down of motivation-specific response patterning. It represents a mismatch of emotion-related behavior to the intentions or “meaning” within a situation. We will also discuss the utility of acoustic and psychophysiological measures in predicting flat affect in schizophrenia, and in differentiating it from depression.