In the face of anger: Startle modulation to graded facial expressions JONATHAN P. DUNNING, ANTHONY AURIEMMO, CLAUDE CASTILLE, and GREG HAJCAK Department of Psychology, Stony Brook University, Stony Brook, New York, USA Abstract In the present study, the startle reflex was examined with respect to the degree of anger displayed in facial expressions. To this end, 52 participants viewed faces that were morphed to display 0, 20, 40, 60, 80, or 100% anger. As the percentage of anger in faces increased from 0 to 100%, faces were perceived as increasingly angry; however, relative to neutral facial expressions, startle amplitude was only potentiated to maximally angry faces. These data imply a non- linear relationship between the intensity of angry faces and defensive physiological activity. This pattern of startle modulation suggests a categorical distinction between threatening (100% anger) and other facial expressions pre- sented. These results are further discussed in terms of existing data, and how this paradigm might be utilized in psychopathology research. Descriptors: Emotion, Startle, Faces, Anger The startle response is a primitive defensive reflex that is observed across species in response to abrupt and intense sensory stimuli (Davis, 1984; Grillon & Baas, 2003). In humans, the eyeblink reflex is one of the most reliable components of the startle re- sponse (Grillon & Baas, 2003; Landis & Hunt, 1939). According to the motivational priming hypothesis, a defensive reflex should be potentiated when the aversive motivational system is active, but should be attenuated when the appetitive motivational sys- tem is active (Lang, Bradley, & Cuthbert, 1997). Indeed, research has supported this notion by demonstrating that eyeblink re- sponses to startle probes are enhanced while participants view unpleasant scenes and attenuated when they view pleasant scenes (Lang, 1995); this pattern holds across various probe (e.g., acoustic, visual, tactile; Bradley, Cuthbert, & Lang, 1990) and affective stimulus modalities (e.g., sounds, movies, odors; Brad- ley, Cuthbert, & Lang, 1999). Extensive research in non-human animals suggests that po- tentiation of the startle reflex by fear-eliciting stimuli depends critically on the amygdala (Davis, Falls, Campeau, & Kim, 1993; Davis, Walker, & Lee, 1999; Lang, Davis, & Ohman, 2000). Studies that employ functional neuroimaging confirm that startle modulation in humans is likewise dependent upon the amygdala (Pissiota, Frans, Michelgard, Appel, Langstrom, et al., 2003). Despite ample evidence that the human startle reflex is potent- iated by complex aversive pictures, fewer studies have assessed whether the startle response is similarly increased while partic- ipants view threatening emotional faces. This is particularly sur- prising in light of the rather large functional neuroimaging literature on the amgydala that has utilized threatening faces as emotional stimuli (Blair, Morris, Frith, Perrett, & Dolan, 1999; Breiter, Etcoff, Whalen, Kennedy, Rauch, et al., 1996; Vuilleumier, Armony, Driver, & Dolan, 2001; Whalen, Rauch, Etcoff, McInerney, Lee, & Jenike, 1998; Whalen, Shin, McInerney, Fischer, Wright, & Rauch, 2001). In one develop- mental study, 5-month-old children had larger startle amplitudes when viewing angry compared to happy or neutral faces (Balaban, 1995; but see Spangler, Emlinger, Meinhardt, & Hamm, 2001). In adult samples, potentiated startle has also been found in the context of angry compared to happy and neutral faces (Hess, Sabourin, & Kleck, 2007; Springer, Rosas, McGetrick, & Bowers, 2007). Thus, angry expressions reflect viewer-directed threat and effectively engage withdrawal or es- cape motivation (Springer et al., 2007). However, one study only found this effect when expressers were male (Hess et al., 2007); and Hess and colleagues suggest that male compared to female faces may indicate more direct threat since men are more asso- ciated with social dominance and aggressive acts than women. Activation of the amygdala depends on the affective strength of the eliciting stimuli, such that amygdala activation increases as ratings of emotional intensity/arousal also increase (Canli, Zhao, Brewer, Gabrieli, & Cahill, 2000; Phan, Taylor, Welsh, Decker, Noll, et al., 2003; Phan, Taylor, Welsh, Ho, Britton, et al., 2004). Data suggests that this type of dose-response relationship may also exist between stimulus aversiveness and potentiation of the startle reflex. For example, Bradley, Codispoti, Cuthbert, and Lang (2001) found potentiated startle responses in the presence of a wide variety of aversive picture content, including contamina- tion, mutilation, and scenes depicting human and animal attack. The authors would like to thank Paul Ekman for providing permis- sion to reproduce Figure 1. Address reprint requests to: Greg Hajcak, Ph.D., Department of Psychology, Stony Brook University, Stony Brook, NY 11794-2500. E-mail: greg.hajcak@stonybrook.edu Psychophysiology, 47 (2010), 874–878. Wiley Periodicals, Inc. Printed in the USA. Copyright r 2010 Society for Psychophysiological Research DOI: 10.1111/j.1469-8986.2010.01007.x 874