Elsevier

Comprehensive Psychiatry

Volume 83, May 2018, Pages 84-88
Comprehensive Psychiatry

Blunted vocal affect and expression is not associated with schizophrenia: A computerized acoustic analysis of speech under ambiguous conditions

https://doi.org/10.1016/j.comppsych.2018.03.009Get rights and content

Highlights

  • Controls and schizophrenia patients exhibited longer pause means and decreased utterances with increased cognitive load.

  • Schizophrenia patients exhibited increased arousal as indicated by increased pitch perturbation and intensity perturbation.

  • These findings suggest that in patients with schizophrenia, vocal affect/alogia is generally unremarkable under ambiguous conditions.

Abstract

Introduction

Patients with schizophrenia are consistently rated by clinicians as having high levels of blunted vocal affect and alogia. However, objective technologies have often failed to substantiate these abnormalities. It could be the case that negative symptoms are context-dependent.

Objectives

The present study examined speech elicited under conditions demonstrated to exacerbate thought disorder.

Methods

The Rorschach Test was administered to 36 outpatients with schizophrenia and 25 nonpatient controls. Replies to separate “perceptual” and “memory” phases were analyzed using validated acoustic analytic methods.

Results

Compared to nonpatient controls, schizophrenia patients did not display abnormal speech expression on objective measure of blunted vocal affect or alogia. Moreover, clinical ratings of negative symptoms were not significantly correlated with objective measures.

Conclusions

These findings suggest that in patients with schizophrenia, vocal affect/alogia is generally unremarkable under ambiguous conditions. Clarifying the nature of blunted vocal affect and alogia, and how objective measures correspond to what clinicians attend to when making clinical ratings are important directions for future research.

Introduction

Schizophrenia is a psychological disorder that severely affects an individual's cognitive, perceptual, and motor functioning and is manifested in positive, negative and disorganized symptoms. Positive symptoms include delusions (i.e. firmly held beliefs that are unchangeable in the presence of incongruent evidence) and hallucinations (i.e. perceptions occurring in the absence of an external stimulus). Negative symptoms include speech-related behaviors such as blunted vocal affect (i.e., reduced vocal variability), and alogia (i.e., diminished vocal production) [1]. Negative symptoms in schizophrenia are a pernicious category of symptoms due to their chronic nature, resistance to medication, and indication of poor prognosis [2]. Additionally, patients with schizophrenia and other psychotic disorders demonstrate difficulty in discerning emotional affect in the speech of other individuals [3,4,5]. Previous research suggests these difficulties may be moderately related to cognitive deficits or impaired theory of mind abilities in individuals with schizophrenia [6].

Patients show profound deficits in speech affect and expression on clinician-rated measures like the Scale for the Assessment of Negative Symptoms (SANS) [7]. A recent review found that ratings of speech variability and production from the SANS were 3–5 standard deviations abnormal compared to nonpsychiatric controls [8]. Importantly, these abnormalities are seen for patients as a group, and not just in subgroups of patients with particularly pronounced negative symptoms [2]. While clinician-rating measures demonstrate acceptable reliability and, in some studies, convergence with a range of clinical phenomenon [2], they are limited in many respects. Of note, they are based on relatively ambiguous operational definitions covering a wide range of behaviors that are not well operationalized or grounded in speech sciences; they lack consideration to cultural, individual difference, and environmental factors that heavily influence speech expression, and offer limited response options that are insensitive to all but major changes over time [9,10,11]. However, technological developments allow for more reliable and precise measures of blunted affect and alogia based on computerized/objective acoustic analysis [10]. These measures offer potentially greater reliability, validity, and sensitivity, leading to greater accuracy in assessment [12]. To date, there have been several dozen studies employing acoustic analysis of patient speech in some capacity. While there are promising “proof of concept” studies (e.g., [13]), the larger literature provides results that are underwhelming in magnitude and consistency. For example, a large multi-site analysis of patient speech found no statistical difference between 309 schizophrenia patients and 117 nonpsychiatric controls on a range of empirically-derived objectively-based vocal measures [14]. Moreover, a meta-analysis of 13 published studies found that, at most, certain measures related to alogia (e.g., pause times/numbers) are exaggerated in some studies on the order of one standard deviation above controls [9,15] and are normal in other studies [8,16]. This is far below the 3–5 standard deviations found from clinical rating scales (see Fig. 1 for a summary). Importantly, clinical ratings of blunted vocal affect often show statistically nonsignificant correlations (−0.18 [9]) to small (0.26 [17]) with acoustically-derived measures of vocal affect (e.g., intonation, emphasis) obtained during laboratory speaking tasks. However, negative and positive symptoms have been shown to be significantly associated with vocal production (e.g., increased pause times and decreased number of utterances) and positive symptoms were correlated with increased inflection [12].

It could be the case that speech production (tapping alogia) and variability (tapping blunted affect) are “context-dependent,” that is, their emergence is contingent upon cognitive, social, or other factors. For example, it has been proposed that cognitive resources related to “on-line” attentional and working memory abilities are particularly important for producing speech [19,20]. Given that these resources are deficient in schizophrenia [21], a cognitive resource model may explain the symptoms of blunted affect and alogia. In support of this idea, increasing experimentally-manipulated cognitive demands causes decreased speech production and variability in schizophrenia patients [19,21]. This could be due to the fact that allocation of cognitive resources for one task limits resources for speech; and speech becomes quantitatively and qualitatively sparser/flatter.

One approach to understanding cognitive and other contextual effects on speech involves the Rorschach inkblot test, a popular measure of positive thought disorder [22]. The Rorschach is thought to be a useful measure of thought disorder because it imposes a relatively heavy “processing” load, vis a vis the “ambiguity” of the inkblot stimuli, which potentially exacerbates thought disorder and, behaviorally, speech incoherence [23]. Although this processing load is potentially imposed via multiple systems (e.g., cognitive, affective, social), analysis of speech from Rorschach tests offers an opportunity to understand whether blunted affect/alogia are exacerbated in the same manner as positive thought disorder. The Rorschach administration utilized in this study involved three distinct phases per card. The first “Perceptual” phase involved participants describing the stimulus in terms of identifiable objects while viewing the card. The second “Memory” phase involved participants describing the features of the stimulus that provoked the aforementioned object without viewing the card. The third “Identify” phase involved participants physically indicating (i.e., pointing to) which areas of the stimulus corresponded with the features described in the “Memory” phase. We consider the second phase to be more taxing than the first because it involves communicating complex semantic material (e.g., explaining and defending their response). Therefore, the “Memory” phase requires the participant to exert more cognitive resources than the “Perceptual” phase.

The current study used computerized acoustic analysis to examine responses from a Rorschach inkblot test. Across phases of the Rorschach, we hypothesized that patients with schizophrenia would exhibit abnormalities in vocal production and vocal expression, as measured by a decrease in number of utterances and increase in pause times, and variability, as measured by flattened intonation, pitch perturbation, emphasis, and intensity perturbation, compared to nonpsychiatric controls. Additionally, we hypothesized that these abnormalities would be exacerbated during the “Perceptual” phase of the Rorschach, which is potentially more taxing in cognitive, motivational, and other abilities than the “Memory” phase. Finally, we examined the relationships between clinical ratings and objectively-defined acoustic variables.

Section snippets

Participants

Participants consisted of 36 outpatients with DSM-IV diagnosed schizophrenia and 25 controls. All participants were recruited as part of a larger study and went through an informed consent process. Portions of the data have been reported elsewhere [24]. Education level and sex didn't differ between groups, though the control group was significantly younger than the schizophrenia group (Table 1). Exclusion criteria for both the schizophrenia and control group included the following: lack of

Results

Women showed significantly more utterances and intonation than men in both conditions (p's < 0.05; across groups). Education level was positively associated with more utterances (p < 0.001). No other demographic or clinical variables were significantly related to the speech variables.

There was a phase effect for Pause Mean and Utterance Number, in that all participants, regardless of diagnostic category, exhibited fewer utterances and shorter pause times in the memory condition as opposed to

Discussion

The current findings indicate that individuals with schizophrenia exhibited largely normal speech, regardless of task. Congruent with prior studies that directly manipulated the speaking task, vocal variables varied in both groups as a function of processing load [19,21]; though this was in the opposite direction than expected. It is unclear why the memory phase produced speech that was more prodigious than in the perceptual phase, but it is likely that contextual factors other than cognitive

Conclusions

In the current study, patients with schizophrenia were not abnormal in vocal expression, even in a context that was putatively cognitively taxing. This study contributes to others [8,12,17], which collectively, present an important question for negative symptoms research in schizophrenia: “When do patients with schizophrenia demonstrate the dramatic vocal deficits captured in clinical ratings”. Of note, future studies might attempt to simulate conditions under which these vocal deficits have

Acknowledgements

The authors wish to thank Deborah L Levy for sharing her data with us.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References (31)

Cited by (5)

  • Investigating the diagnostic utility of speech patterns in schizophrenia and their symptom associations

    2021, Schizophrenia Research
    Citation Excerpt :

    This method was chosen to mimic a more natural speaking environment, as opposed to other methods such as picture descriptions (Çokal et al., 2018), or dream reports (Mota et al., 2017). Previous work suggests speech volume varies by elicitation scenarios (Meaux et al., 2018). The SZ patients may have also demonstrated inefficient speech production in this scenario, requiring more words, utterances and turns than HCs to convey their intended meanings and responses.

  • Voice patterns in schizophrenia: A systematic review and Bayesian meta-analysis

    2020, Schizophrenia Research
    Citation Excerpt :

    On the contrary, voice atypicalities due to differences in fine motor control of the vocal folds should be similar across tasks. Further, a few studies have adopted a more fine-grained perspective, and assessed the relationship between acoustic measures and clinical features with some promise; however, the findings are still very sparse (Alpert et al., 2002; Alpert et al., 2000; Cohen et al., 2016; Meaux et al., 2018; Zhang et al., 2016). Finally, more recent studies have tried to capitalize on the technological advancements in speech signal processing, and the application of multivariate ML techniques to better capture the complex, multivariate and often non-linear nature of acoustic patterns (Bone et al., 2017; Huys et al., 2016; for an introduction to ML techniques in the context of voice analysis see also the appendix to Fusaroli et al., 2017).

Submitted as a “Research Paper” to Comprehensive Psychiatry.

View full text