Recognition and evaluation of emotional prosody in individuals with generalized social phobia: A pilot study

Abstract

Studies using facial emotional expressions as stimuli partially support the assumption of biased processing of social signals in social phobia. This pilot study explored for the first time whether individuals with social phobia display a processing bias towards emotional prosody. Fifteen individuals with generalized social phobia and fifteen healthy controls (HC) matched for gender, age, and education completed a recognition test consisting of meaningless utterances spoken in a neutral, angry, sad, fearful, disgusted or happy tone of voice. Participants also evaluated the stimuli with regard to valence and arousal. While these ratings did not differ significantly between groups, analysis of the recognition test revealed enhanced identification of sad and fearful voices and decreased identification of happy voices in individuals with social phobia compared with HC. The two groups did not differ in their processing of neutral, disgust, and anger prosody.

Introduction

Etiological models as well as an extensive body of research suggest that the maintenance of social phobia might be due to cognitive biases in the processing of social information (e.g. Clark & Wells, 1995; Heinrichs & Hofmann, 2001). In particular, social anxiety has been associated with an increased sensitivity towards signals of negative evaluation by others, even when these signals are ambiguous (Stopa & Clark, 2000; Yoon & Zinbarg, 2007). Consistent with this notion, hypersensitivity to negative social signals in social phobia has been demonstrated across studies of attention, memory, interpretation, and/or judgment (for a review, see Heinrichs & Hofmann, 2001), by means of changed explicit emotional evaluation of stimuli (Straube, Kolassa, Mentzel, & Miltner, 2004; Straube, Mentzel, & Miltner, 2005), and/or by increased brain activation (e.g. Amir et al., 2005; Phan, Fitzgerald, Nathan, & Tancer, 2006; Stein, Goldin, Sareen, Zorrilla, & Brown, 2002; Straube, Kolassa, Mentzel, & Miltner (2004), Straube, Mentzel, & Miltner (2005)).

Importantly, the majority of studies on social phobia have utilized emotional facial expressions (EFE) as stimuli conveying negative social signals. For example, functional neuroimaging studies showing increased brain activation to facial expressions of anger, contempt, disgust, and fear in socially phobic compared with nonphobic individuals support the assumption of enhanced processing of negative social signals in social phobia (Amir et al., 2005; Phan et al., 2006; Stein et al., 2002; Straube, Kolassa, Mentzel, & Miltner (2004), Straube, Mentzel, & Miltner (2005)). Behavioral studies also indicate that social phobic individuals show increased attentional (Eastwood et al., 2005; Mogg, Philippot, & Bradley, 2004; Pishyar, Harris, & Menzies, 2004) and biased memory processing (Coles & Heimberg, 2005; Foa, Gilboa-Schechtman, Amir, & Freshman, 2000) of negative EFE (but see Chen, Ehlers, Clark, & Mansell, 2002; Pérez-López & Woody, 2001, for an opposing view) in contrast to nonphobic controls. Furthermore, enhanced recognition of negative facial cues in socially anxious compared to less anxious controls has been observed (Joormann & Gotlib, 2006; Winton, Clark, & Edelmann, 1995; but see Philippot & Douilliez, 2005). Remarkably though, most studies have failed to find any evidence that individuals with social phobia differ from normal controls in the explicit evaluation of negative EFE on dimensions such as valence or arousal (Amir et al., 2005; Merckelbach, Van Hout, Van den Hout, & Mersch, 1989; Philippot & Douilliez, 2005; Stein et al., 2002; but see Straube, Kolassa, Mentzel, & Miltner (2004), Straube, Mentzel, & Miltner (2005)). Moreover, findings obtained with negative EFE have recently been complemented by studies showing that positive EFE can also elicit biased processing in social phobia (Phan et al., 2006; Straube et al., 2005). Although socially phobic individuals seem to evaluate happy facial expressions as pleasant (Philippot & Douilliez, 2005) and even more arousing than nonphobic controls (Straube et al., 2005), they take longer to recognize happy EFE than normal controls (Silvia, Allan, Beauchamp, Maschauer, & Workman, 2006), and lack a positivity bias towards happy faces shown by healthy individuals (Coles & Heimberg, 2005, for memory bias; Eastwood et al., 2005; Pishyar et al., 2004, for attentional bias).

Surprisingly, assumptions about cognitive biases towards the processing of social cues in social phobia have hardly been investigated with expressive social signals other than facial expressions. One should expect, however, to see disorder-associated cognitive biases towards such cues across a range of stimuli and modalities. For example, it is unknown whether a biased processing of emotional signals in social phobia is also displayed in response to prosodic components of human speech. Due to typical acoustic patterns based on pitch contour, fundamental frequency, tempo, loudness, and stress, a reliable recognition of emotions from speech can be accomplished (Banse & Scherer, 1996). In consequence, prosodic features can reveal a speaker's emotional state and intention towards others and provide powerful signals of social approval and disapproval. Recently, it has been emphasized that the recognition of emotional prosody is impaired in various neurological and psychiatric disorders (see Bostanov & Kotchoubey, 2004). Psychopathic individuals, for example, have been observed to be restricted in their recognition of fearful emotional prosody compared with healthy controls (HC) (Blair et al., 2002). Depressive individuals, in contrast, have been shown to suffer from a general impairment of prosody recognition across different types of emotional categories (Luck & Dowrick, 2004). Although of high relevance, the response of individuals with social phobia towards cues of emotional prosody has not yet been the focus of investigations. Rather, a review article on cognitive biases in social phobia has emphasized the fact that studies on auditory processing biases in social phobia are seriously lacking (Heinrichs & Hofmann, 2001). Thus, it remains to be examined whether attentional, mnemonic, and recognition biases, or changed explicit valence and arousal ratings towards emotional prosody may be found in individuals with social phobia compared with HC.

Therefore, this pilot study aimed to explore the processing of emotional prosody in individuals with social phobia. Specifically, we investigated whether individuals suffering from generalized social phobia (GSP) display a recognition bias for emotional voices compared with HC. According to previous work on the presence of a recognition bias for facial expressions in GSP, we examined whether individuals with GSP would show improved recognition of negative prosodic stimuli. Based on current theories proposing an increased sensitivity for social signals of rejection and disapproval in social phobia in particular, we expected that such a recognition bias should be found especially for anger prosody. Furthermore, we explored whether individuals with GSP would evaluate emotional prosody differently with regard to arousal and valence as compared with HC. We predicted that individuals with GSP would experience negative—especially anger—prosody as more negative and more arousing than HC.