Emotional modulation of the postauricular reflex

More documents

Recommendations

Info

Emotion and the post-auricular reflex 429 Procedure Participants provided informed written consent and then completed a biographical form that screened for physical ailments, medication use, and visual and auditory impairments. Following electrode attachment, participants were advised they would be viewing a series of pictures and rating their reactions to each. They were instructed to watch each picture the entire time it was on the screen and disregard occasional noises occurring through the headphones or earphones. Before the main picture series began, participants were given a demonstration of the Self- Assessment Manikin (SAM; Lang, 1980) rating procedure, in which they characterized their reactions to pictures on dimensions of valence, arousal, and dominance (cf. Lang, 1980), as well as interest (cf. Levenston et al., 2000). Each of the 66 pictures was preceded by a blank screen for 3 s and was then presented for 6 s. Seven seconds after picture offset, the ratings display appeared and the participant completed the four ratings. The interval between completion of the ratings and the onset of the next picture ranged between 8 s and 14 s, averaging 11 s. Data Processing and Analysis Because the post-auricular reflex is a ‘‘microreflex’’ with a low signal-to-noise ratio (cf. Hackley et al., 1987), averaging of trials within condition was required to score the reflex. For one set of analyses, we aggregated the rectified post-auricular waveforms by thematic content within picture valence category. Because there were equivalent numbers of human and object pictures within the neutral category, we aggregated across trials within each of these neutral contents, and we also aggregated across trials for erotic and adventure contents within the pleasant category, and victim and threat contents within the unpleasant category. Each aggregate waveform thus incorporated data for nine picture trials. A nine-trial aggregate waveform incorporating data for all ITI probes was also formed. Post-auricular reflex peak magnitude was scored from these aggregate waveforms, along with mean post-auricular muscle activity during the preprobe baseline. Effects of picture valence were examined by collapsing magnitude scores across contents within valence category. A subsidiary set of analyses was performed to examine effects of affective stimulus intensity on post-auricular reflex magnitude. For these analyses, we aggregated post-auricular waveforms by picture intensity level (low, high) within pleasant and unpleasant valence categories, again yielding aggregate waveforms consisting of nine trials per condition, and extracted peak magnitude scores from these aggregates. Using the resultant scores, the following analyses were performed: 1. To examine the effect of visual foreground engagement on post-auricular reflex magnitude, we conducted a one-way repeated measures ANOVA in which the neutral picture condition was compared against the ITI (no-picture) condition. A parallel one-way ANOVA was conducted to examine the effect of foreground engagement on post-auricular baseline activity. 2. To examine effects of picture valence (pleasant, neutral, unpleasant) and ear recording site (left, right) on postauricular magnitude, a 3 2 multivariate ANOVA was performed with these variables included as within-subjects factors. Significant main effects were subsequently examined in terms of orthogonal linear (e.g., pleasant vs. unpleasant) and quadratic (e.g., pleasant/unpleasant vs. neutral) contrasts (cf. Bradley, Cuthbert, & Lang, 1993; Bonnet et al., 1995; Moulder et al., 1995). A supplemental one-way MANOVA was performed to examine effects of picture valence on baseline post-auricular activity. 3. With regard to picture content and intensity, we were interested in whether affect-modulation effects for the postauricular reflex would be tied to specific contents or to pictures of high intensity. We addressed these questions by performing two sets of planned contrasts using one-way ANOVAs. One involved comparisons of affective contents within each valence against each other, and the other entailed contrasts of pleasant and unpleasant pictures of each intensity (low, high) against each other. An a level of .05 was used as the criterion for significance in all statistical tests. Results Picture Foreground versus ITI Post-auricular reflex magnitudes were smaller during neutral pictures compared with ITIs, Ms(SDs) 5 6.64 (6.18) mVand 7.98 (5.71) mV, respectively, F(1,23) 5 4.65, p 5 .04, Z 2 5 .17. There was also a trend for baseline post-auricular EMG activity during neutral pictures to be smaller than during ITIs, Ms(SDs) 5 0.74 (0.32) mV and 0.84 (0.36) mV, respectively, F(1,23) 5 2.92, p 5 .10, Z 2 5 .11. Picture Valence Figure 1 displays grand average waveforms for the post-auricular reflex by picture valence, with average reflex magnitudes depicted in the inset bar graph. A significant main effect of picture valence was found for post-auricular reflex magnitude, F(2,22) 5 4.16, p 5 .03, Z 2 5 .27. 4 As shown in Figure 1, the post-auricular reflex was larger during pleasant pictures than unpleasant pictures, linear Valence F(1,23) 5 8.50, p 5 .01, Z 2 5 .27. Follow-up paired t tests showed that post-auricular magnitudes tended to be greater during pleasant than neutral pictures, t(23) 5 1.94, p 5 .07, whereas post-auricular magnitudes during neutral and unpleasant pictures did not differ, t(23) 5 1.11, p 5 .28. No main effect of ear recording site (right vs. left) was found, F(1,23) 5 1.00, p 5 .33, Z 2 5 .04, and the Ear Valence interaction was not significant, F(2,22) 5 0.31, p 5 .74, Z 2 5 .01. As further evidence of the consistency of the post-auricular valence modulation effect, 75% (18/24; 95% confidence interval 5 55%–88%) of participants showed numerically greater post-auricular reflex magnitudes during pleasant compared with unpleasant pictures, and 71% (17/24; 95% confidence interval 5 51%–85%) showed numerically greater post-auricular reflex magnitudes during pleasant versus neutral pictures. Both proportions significantly exceeded chance, according to formulae provided by Newcombe (1998). The MANOVA with picture valence as the within-subjects factor and mean pre-probe baseline post-auricular EMG activity 4 A supplementary analysis was performed in which picture trials were subdivided into three sequential blocks, and post-auricular waveforms were aggregated by valence (pleasant, neutral, unpleasant) within each block. As in the primary MANOVA, this analysis yielded a significant valence main effect, and also a significant main effect for block, F(2,22) 5 6.77, p 5 .011, Z 2 5 .338, but no Block Valence interaction, F(4,20) 5 0.95, p 5 .441, Z 2 5 .040. Thus, while the magnitude of the post-auricular reflex declined over the session, the emotional modulation of the post-auricular reflex remained constantFconsistent with the findings of Bradley et al. (1993) for the startle blink reflex.
430 S.D. Benning, C.J. Patrick, and A.R. Lang PA EMG Activity (µV) 6 5 4 3 2 1 0 -25 0 25 50 75 100 125 as the dependent variable revealed no effect of picture valence, F(2,22) 5 0.30, p 5 .74, Z 2 5 .03. Post-auricular baseline Ms (SDs) for pleasant, neutral, and unpleasant pictures were as follows: 0.74 (0.29) mV, 0.75 (0.32) mV, and 0.72 (0.34) mV, respectively. Affective Content and Intensity Within pleasant contents, the magnitude of post-auricular reflexes during erotic pictures did not differ from those during adventure pictures, Ms 5 7.23 and 7.21 mV (SDs 5 5.86 and 6.12), respectively, F(1,23) 5 0.00, p 5 .96, Z 2 5 .00. Likewise, within unpleasant contents, post-auricular reflex magnitudes did not differ between victim and threat pictures, Ms 5 6.24 and 6.38 mV(SDs 5 5.20 and 5.58), respectively, F(1,23) 5 0.11, p 5 .75, Z 2 5 .01. Post-auricular reflex magnitudes were significantly greater during high-intensity pleasant pictures than during high-intensity unpleasant pictures, Ms 5 8.11 and 6.81 mV (SDs 5 6.87 and 5.59), respectively, F(1,23) 5 7.78, p 5 .01, Z 2 5 .25, but postauricular reflex magnitudes during low-intensity pleasant and unpleasant pictures did not differ significantly from each other, Ms 5 7.77 and 7.18 mV (SDs 5 6.48 and 6.18), respectively, F(1,23) 5 2.31, p 5 .14, Z 2 5 .09. Discussion Pleasant Neutral Unpleasant PA Peak Magnitude (µV) 8 7 6 5 Time Relative to Probe Onset (ms) Pleasant Neutral Unpleasant Picture Valence Figure 1. Grand average post-auricular (PA) reflex waveforms (main panel) and magnitudes SE (inset) by picture valence. Our findings indicate that the magnitude of the post-auricular reflex is reliably modulated by emotional valence. Specifically, we found that the post-auricular reflex tended to be potentiated during viewing of pleasant pictures compared with neutral, and inhibited during viewing of aversive pictures, particularly during high-intensity pictures. 5 This linear modulation pattern is 5 In a previous investigation of affect and attention during picture viewing that employed a wider range of probe times (Bradley, Drobes, & Lang, 1996), measurement of the post-auricular reflex was included along with assessment of blink startle and tone probe responses (M. M. Bradley, pers. comm., May 7, 2003). Although post-auricular reflex magnitude in this study was somewhat higher during pleasant versus unpleasant pictures, the difference was not significant. On the other hand, more recently, a significant linear valence effect on post-auricular reflex magnitude, mirroring the finding reported here, was obtained by investigators at Indiana University (Ames, Merritt, Stout, & Hetrick, 2003). opposite to that observed for the defensive startle blink reflex (Lang et al., 1990). Analyses of post-auricular muscle activity during the pre-probe baseline period indicated that this modulatory effect was not attributable to differences in baseline EMG (as would be expected if these results were due to head motion or eye gaze; Patuzzi & O’Beirne, 1999; Stekelenburg & van Boxtel, 2001, 2002), because baseline post-auricular activity did not differ across picture valence categories. On the other hand, baseline post-auricular activity was reduced in connection with attenuated post-auricular reflex response during viewing of neutral foregrounds in relation to ITIs. This finding is consistent with the idea that tonic activity in various somatic systems, including the postauricular musculature, decreases to enhance perceptual sensitivity during orienting (Stekelenburg & van Boxtel, 2001, 2002). Lang et al. (1990) proposed a motivational priming hypothesis to account for modulatory effects of emotional valence on the startle reflex. Citing Konorski (1967), they postulated that defensive reflexes like the startle blink are primed during negative emotional states and inhibited during positive emotional states. The potentiation effects observed for the neck (Cassella, Harty, & Davis, 1986), blink (Lang et al., 1990), and whole-body (Davis et al., 1982) reactions to noise probes under conditions of threat are consistent with the idea that these are components of a common defensive startle response. Although the post-auricular reflex is likewise elicited by a sudden, loud noise, our data argue against the post-auricular reflex being a short latency component of the defensive startle reflex (Hackley, 1993). Indeed, there is evidence that the post-auricular reflex may have a different circuitry than the startle reflex (Cassella & Davis, 1986; Hackley, 1993), though more recent research with specific excitotoxic lesions in the circuitry of the pinna reflex in rats (rather than with general fiber-destroying electrolytic lesions; Cassella & Davis, 1986) suggests that the circuitry for the post-auricular reflex may also include the nucleus reticularis pontis caudalis (Davis, Walker, & Yee, 1999; Li & Frost, 1996). Lang et al.’s (1990) motivational priming hypothesis postulates that appetitive reflexes (salivation) would be facilitated during positive emotional states and inhibited during negative emotional states. In this regard, the post-auricular reflex behaved as an appetitive reflex in the current study, as it was potentiated during viewing of pleasant pictures and attenuated during viewing of unpleasant pictures. The fact that the post-auricular reflex increased during pleasant picture viewing indicates that some other mechanism (appetitive motor priming) operated against the influence of cross-modal sensory engagementFanalogous to the impact of defense system activation on the startle reflex during unpleasant pictures, which opposes the inhibitory influence of cross-modal attention in picture viewing (Cuthbert et al., 1996; Lang et al., 1997). In the case of the startle reflex, a specific neural mechanism has been identified to account for the facilitatory effect of aversive foreground stimulationFnamely, an input from the amygdala to the startle reflex circuit (Davis et al., 1982). A corresponding input from the brain’s appetitive system to the post-auricular circuit could conceivably account for potentiation of the post-auricular reflex during pleasant foreground stimulation. Research on the neural circuitry of the pinna reflex, the animal analogue of the post-auricular reflex, indicates that the motoneurons of the facial nerve that innervates the pinna receive additional input from the retrorubral nucleus (Li & Frost, 1996), a midbrain dopaminergic structure that has been associated with sensitivity to reward stimuli in rats (Waraczynski & Perkins,
Page 1 and 2: Psychophysiology, 41 (2004), 426-43
Page 3: 428 S.D. Benning, C.J. Patrick, and
Page 7: 432 S.D. Benning, C.J. Patrick, and

Emotional modulation of the postauricular reflex

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?