CLINICAL Research Report Startle as a physiological measure of regulation Brittany C. Speed1 1Department of Psychology, University of British Columbia

Edited by: Virginie Cousineau, Department of Psychology, University of British Columbia. Received for review January 5, 2012, and accepted February 10, 2012. Copyright: © 2012 Speed. This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivates License, which permits anyone to download and share the article, provided the original author(s) are cited and the works are not used for commercial purposes or altered in any way.

Abstract In the present study, the startle blink reflex is used as a measure of emotion regulation to affective picture stimuli. Based on the biphasic theory of emotion, it is hypothesized that the startle response will be largest in magnitude in the presence of negative emotional stimuli (Vrana, Spence, & Lang, 1988). It is also hypothesized that when attempting to decrease emotion, participants will show smaller blink magnitudes to negative images, and larger blink magnitudes when attempting to decrease emotion to positive images due to the aversive nature of the startle reflex. The present study highlights the difficulty of finding emotion regulation to positive images with the startle blink paradigm. Participants were 6 female undergraduate students who viewed negative, positive and neutral affective picture stimuli, and attempted to either maintain or suppress their emotional responses to the images. Emotion modulation of the startle response was recorded before regulation instruction onset. Significant differences were found in blink magnitude for emotion modulation between positive and negative images, but neutral images were not significantly different from either. For negative images, blink magnitudes during emotion suppression were significantly smaller than when maintaining emotion. No significant regulation differences were found for the positive images. Keywords: psychophysiology, emotion regulation, emotion modulation, startle blink

Emotion regulation can be thought of as depression (Davidson, 2000; Jackson, Larson, consciously or unconsciously altering one's & Davidson, 2000), and eating disorders behavioural, cognitive, and/or physiological (Clyne, Latner, Gleaves, & Blampied, 2010; emotional response tendencies towards a Harrison, Sullivan, Tchanturia, & Treasure, goal (Gross, 1998). The ability to regulate 2010; Whiteside, Chen, Neighbors, Hunter, is important for both mental and Lo, & Larimer, 2007). Understanding possible physical well-being, and social interaction mechanisms associated with emotion (Kim & Hamann, 2007). Dysfunction in regulation difficulties are essential to emotion regulatory abilities has been addressing these problems in clinical thought to play a role in many forms of populations. Although emotion regulation psychopathology, such as anxiety, deficits are a core feature of many forms of

Corresponding Author: Brittany C. Speed E-mail: [email protected]

CLINICAL Speed Research Report

psychopathology, physiological evidence of attenuated, or smaller in magnitude, due to these emotion regulatory deficits is severely a mismatch between the reflex and the lacking. A large body of research that has emotional context (Lang et al., 1990; Vrana been used to correlate psychopathology and et al., 1988). Indeed, research has shown emotion regulation has been self-report, that the startle blink reflex is an effective (Aldao & Nolen-Hoeksema, 2010; Whiteside measure of emotion modulation, with the et al., 2007) and the ability of individuals to startle reflex being potentiated in the accurately assess their own emotion presence of negative emotions, attenuated regulatory abilities has not been confirmed. in the presence of positive emotions, and Little research exists that overtly examines moderate in size when there is a lack of emotion regulation and its possible deficits emotion (Bradley, Cuthbert, & Lang, 1991; in an experimental setting using Vrana et al., 1988). An increase in startle psychophysiological measures of emotion magnitude from positive, neutral, to regulation. negative emotional contexts has been found The startle blink paradigm is an when emotions are elicited using visual advantageous physiological tool as it has the stimuli (Bradley et al., 1991; Vrana et al., ability to differentiate between positive and 1988), olfactory stimuli (Miltner, Matjak, negative emotional states. This paradigm Braun, Diekmann, & Brody, 1994), narrative involves administering a loud blast of white stimuli, such as remembering events (Cook, noise, which acts to startle the participant Hawk, Davis, & Stevenson, 1991), and threat causing their eyes to twitch reflexively. This of shock paradigms (Greenwald, Bradley, startle reflex is measured using Cuthbert, & Lang, 1998; Lissek et al., 2007). electromyographic (EMG) sensors placed on Research showing that the startle blink the obicularis oculi muscle below the left response is an effective measure of emotion eye. The startle reflex triggers the defensive regulation to affective stimuli is less aversive response of the biphasic appetitive- conclusive. The conditions common to aversive dimension of emotion; this emotion regulation research are: “maintain”, automatic response is enhanced when the which instructs participants to focus on their emotional context matches the reflex (Vrana emotional response to the stimuli without et al., 1988). Funayama, Grillon, Davis, & changing it, “suppress”, which involves Phelps (2001) found that the startle decreasing emotional response, and response is mediated by the right medial “enhance”, which involves increasing temporal lobe, specifically the , emotional response to a stimuli (Eippert et which plays a key role in aversive or negative al., 2007; Jackson et al., 2000; Lissek et al., emotional states, supporting the theory that 2007; Ray, McRae, Ochsner, & Gross, 2010). the startle response is an aversive reflex. Jackson et al. (2000) found both emotion Therefore, the startle response is expected modulation and subsequent regulation to to be potentiated, or larger in magnitude, negative pictures compared to neutral. when it occurs in the presence of aversive Replication studies by Lissek et al. (2007), emotional states (Lang, Bradley, & Cuthbert, which included a threat of shock paradigm, 1990; Vrana et al., 1988). Along the same and Lee, Shackman, Jackson, & Davidson reasoning, one would expect that in the (2009) had similar findings of modulation presence of appetitive or positive emotional and regulation to aversive contexts states the startle reflex would be compared to neutral.

UBCUJP – In Press – Volume 1

CLINICAL PHYSIOLOGICAL EMOTION REGULATION Research Report

In concordance with previous research, included positive emotion regulation have Eippert et al. (2007) and Ray et al. (2010) hypothesized that arousal, and not valence, found larger blink magnitudes for the influences the startle response (Dillon & La negative images from the enhance condition Bar, 2005; Driscoll et al., 2008). Dillon & La compared to maintain and suppress Bar (2005) compared regulation to positive, conditions. However, both studies failed to neutral, and negative stimuli and found that find significant startle blink differences for the enhance cue during positive and the suppress condition compared to the negative picture presentation elicited larger maintain condition. One possible reason for blink magnitudes than the maintain and this finding is that the timing of the probes suppress cues, this finding runs counter to may have been too early to accurately what would be expected from the aversive detect suppression, occurring 2 s after the matching hypothesis, where enhancing regulation cue compared to the common 3 s positive emotion should elicit the smallest delay (Eippert et al., 2007). In further blink magnitudes (Bradley et al., 1993; Lang support of this argument, fMRI results et al., 1990; Vrana et al., 1988). The authors showed suppressed amygdala activity during contended that this finding suggests that the the suppress condition compared to the startle blink response is arousal dependent, maintain and enhance conditions, but the meaning that the level of arousal will alter suppressed activation occurred after the the startle magnitude, regardless of startle probe measured regulation. Another emotional valence. Therefore, increasing possible explanation for both studies failing arousal will always result in larger startle to detect significant regulation of the startle responses, whether the emotion is positive response for the suppress compared to the or negative in valence (Dillon & La Bar, maintain cue is that both focused only on 2005). However, significant differences in reappraisal as an emotion regulation blink magnitude for positive compared to technique, as opposed to subjects choosing neutral stimuli was not found, even though their own regulation strategies. Perhaps the positive images were rated as cognitive reappraisal, a cognitive strategy significantly more arousing then the neutral that involves re-evaluating the meaning of images. Therefore these findings are not an emotional (Gross, 1998), is not sufficient to conclude that the startle the most effective negative emotion response is solely impacted by arousal, suppression strategy for all individuals. clearly valence plays some role. Also, unlike For all of the emotion regulation studies Jackson et al. (2000) and others (Lee et al., mentioned thus far, positive picture stimuli 2009; Lissek et al., 2007) no significant was not included. Jackson et al. (2000) difference was found between suppress and contends that the failure to include positive maintain conditions. Possible explanations images was due to a difficultly finding for these contradictory findings may be that attenuation of the startle response to Dillon & La Bar (2005) did not utilize a positive emotion compared to neutral during within-groups design, which is common to a pilot study. Difficulty finding the expected these studies, because of large individual modulation to positive affective stimuli differences in blink magnitudes, combined compared to neutral stimuli is not with the problems associated with using a uncommon (Dillon & La Bar, 2005; Driscoll, small sample size. Tranel, & Anderson, 2008). Studies that have

UBCUJP – In Press – Volume 1

CLINICAL Speed Research Report

Driscoll et al. (2008) also looked at the of the research fails to include positive regulation of positive and negative emotion, affective stimuli (Jackson et al., 2000; Lee et and found that during the enhance al., 2009; Lissek et al, 2007), while others condition, regardless of valence, there were employ different methodologies or have larger blink magnitudes than during the very few participants (Dillon & La Bar, 2005; suppress condition. The authors concluded Driscoll et al., 2009). that these findings suggest that arousal is The purpose of this study is to replicate the main factor that controls the startle and extend the findings of previous research response. However, across regulation on emotion regulation using the startle blink conditions the startle response for negative paradigm (Jackson et al., 2000), with the images was significantly larger than for inclusion of positive affective stimuli. To positive images, even though both were overcome the limitations of previous matched on arousal. Being matched on research a balanced design will be used, arousal and maintaining differences in matching the frequency of picture type and startle blink magnitude suggests that the regulation cues. Probe and trial times that startle response is also influenced by are consistent with previous research of valence. It should also be noted that Driscoll emotion modulation and regulation will also et al. (2008) used a small sample size in their be used (Eippert et al., 2007; Jackson et al., analysis, and consequently their findings 2000). should be interpreted with caution. Emotion modulation is of high Clearly, previous research utilizing importance in this study as it is an physiological measures to assess emotion established finding in startle paradigm regulation yields inconsistent and confusing research (Bradley et al., 1991; Sanchez- findings, and the exact relationship between Navarro et al., 2008; Vrana et al., 1988); the startle response and positive emotional failing to obtain the expected linear cues is not agreed upon or well understood. relationship of valence may suggest Possible reasons for a failure to find startle problems in the paradigm that limit the response differences between neutral and ability to interpret the emotion regulation positive stimuli may be due to a difficulty findings. Subjective ratings of valence and eliciting positive emotion in a lab setting arousal to the images will also be measured (Jackson et al., 2000), and subjective in conjunction with normative ratings in differences in interpreting stimuli as order to assure that arousal between positive. Research shows that arousal also positive and negative images are matched, influences the startle response (Dillon & La and therefore not responsible for Bar, 2005; Driscoll et al., 2008; Lang et al., differences in regulation. 1990). To effectively measure emotional In this study the independent variable is valence, arousal must be balanced for the combined picture types and regulation positive and negative stimuli; however, cues, which result in 6 distinct conditions. some research studies fail to take into The dependent variable is the blink account the participant’s subjective ratings magnitude elicited by the startle paradigm. of arousal or valence to the stimuli (Larson Based on this information two main et al., 2000; Lee et al., 2009). Problems with hypotheses were generated. The first current emotion regulation research using hypothesis is that the startle response will the startle blink paradigm are that a majority show attenuation during the suppress

UBCUJP – In Press – Volume 1

CLINICAL PHYSIOLOGICAL EMOTION REGULATION Research Report negative condition compared to the maintain negative condition. The second Stimuli hypothesis is that the startle response will All images were selected from the IAPS show potentiation during the suppress based on normative female arousal and positive condition compared to the maintain valence ratings. Pictures were selected to positive condition. These hypotheses are create three discrete picture types: negative based on the concept that the startle (low valence, high arousal), positive (high response is a defensive response to threat, valence, high arousal), and neutral (medium activating the aversive domain of the valence, low arousal). Thirty-two of each appetitive-aversive dimension of emotion picture type (positive, negative and neutral) (Vrana et al., 1988). Therefore, when a were included in the picture set, 16 per negative emotional state is suppressed, the block. The overall normative female ratings magnitude of the startle reflex will be for valence were negative M = 2.16, SD = decreased, and when a positive emotional 0.61, positive M = 7.56, SD = 0.61, and state is suppressed the startle reflex will neutral M = 5.0, SD = 0.23. Paired t-tests increase in magnitude. revealed significant differences for valence between negative and positive images t(62) Method = -35.16, p< 0.001, negative and neutral images t(62) = -24.5, p< 0.001, and positive Participants and neutral images t(62) = -22.06, p< 0.001. All procedures were approved by the The overall female normative arousal ratings University of British Columbia Behavioural were negative M = 6.47, SD = 0.74, positive Research Ethics Board, and all participants M = 6.02, SD = 0.76, and neutral M = 2.94, provided informed consent before SD = 0.5. Paired t-tests revealed significant participation. Participants were eight female differences for arousal between negative undergraduate students from the University and neutral images t(62) = 22.51, p< 0.001, of British Columbia. This study included only positive and neutral images t(62) = -19.12, females as it was part of a pilot test for a p< 0.001, and positive and negative images larger study involving emotion regulation t(62) = 2.42, p< 0.05. However, follow-up t- and disordered eating behaviours. As tests between negative and positive images disordered eating is much more common for all probed conditions revealed no among females, males were excluded. Three significant differences in arousal between participants were recruited through the the positive and negative conditions. Each University’s Human Subject Pool for 1.5 block was organized in quasi-random order, course credits; the remaining five ensuring that the pictures were participants were volunteers from the counterbalanced for regulation instruction, University’s Clinical and Cognitive order of presentation, and time of startle Neuroscience Lab. Participants ranged in age probe. No more than three picture types from 20 to 23 years (M = 22, SD = 1.06), and (same valence, regulation cue, or probe all participants had a minimum of 10 years time) occurred in a row. of English fluency. Two participants were excluded from analysis due to a failure to Procedure stay awake and alert during picture Each participant came in for a testing session presentation. that lasted approximately 1.5 hrs. Details

UBCUJP – In Press – Volume 1

CLINICAL Speed Research Report

were then provided about the nature of the presented 4 times during the trial to task, and each participant provided decrease the predictability of the probes. informed consent. EMG sensors were then There were 6 conditions in total, combining applied to the obicularis oculi muscle of the each picture type and regulation cue, which left eye. EMG sensors at the obicularis oculi each occurred 8 times during the course of site are used to measure the size of the the experiment. No probe was presented for startle response, as this muscle is an 24 of the trials. After the task was completed important component of the startle reflex. the EEG and EMG sensors were removed Electroencephalographic (EEG) sensors were and the participants filled out an image also applied at this time as part of another rating for each picture on arousal and study; the EEG data will not be included in valence using the Self Assessment Manikin the analysis. (SAM) valence and arousal scales, with 9 After receiving both visual and verbal rating high (arousal or valence) and 1 rating instructions and engaging in a practice low, this is the same scale that was used to session, participants viewed digitized colour measure the normative ratings of the IAPS images from the International Affective (Bradley & Lang, 1994). Pictures were Picture Set (IAPS; Lang et al., 1999). Each presented in the same order that they had picture was presented for 8 s. A total of 96 originally appeared and picture-viewing time pictures were presented in 2 blocks of 48 was recorded to assess interest. Participants pictures each. At 4 s post-stimulus onset the were also asked to fill-out a strategy regulation cue, either a white equal sign or a questionnaire in which they described the red minus sign, appeared on the screen strategies they used to regulate their briefly. A white equal sign instructed emotions for each picture type (decrease participants to maintain their emotional and maintain for positive, negative and response to the picture. A red minus sign neutral images). After the strategy instructed participants to decrease the questionnaire was complete all participants intensity of the emotion they were were debriefed on the nature of the study. experiencing in response to the picture. Both regulation instructions were presented for Emotion regulation instructions all picture types (positive, negative, and Participants were left free to decide on how neutral) an equal number of times to to regulate their emotions effectively. No maintain consistency across trials. After 8 s a strategies were provided on how to regulate black screen replaced the picture for 5 s. The emotions. However, to separate emotion word “RELAX” then appeared on the screen regulation from inattention, participants for 5 s, at this time participants were were instructed to focus on the picture, and instructed to stop decreasing or maintaining told not to produce thoughts or images that their emotional response to the previous were unrelated to the picture or emotion picture and to get ready for the next picture. they were experiencing. For example, in the During picture presentation an acoustic sample trial participants are shown a picture startle probe (a 95 dB, 50 ms burst of white of a striking snake. Participants are then told noise generated by the Audacity 1.3 Beta that if they are presented with a red minus Unicode Software) was presented to both sign and are required to decrease their ears at either 3 s (probe A) or 7 s (probe B) emotion, is the example given, they after picture onset. Inter-trial probes were should accomplish this by decreasing the

UBCUJP – In Press – Volume 1

CLINICAL PHYSIOLOGICAL EMOTION REGULATION Research Report intensity of the fear they are experiencing, noise in the baseline period (>10 µV), if and not by thinking of something unrelated there was a visible artifact in the segment, to the picture or by trying to produce a or if the magnitude was greater than 3 different emotion. A similar example is standard deviations above the subject’s provided for a positive picture in the individual mean amplitude. Trials with no maintain regulation instruction. All discernable startle response were given a participants are also given a 9 trial, 3 of each magnitude of zero and included in the picture type, practice session and question analysis (Larson et al., 2005). Startle period before the actual experiment begins responses were standardized using within to ensure that they understood the subject z-score conversions to normalize regulation instructions. data, and to reduce the influence of between subjects variability unrelated to EMG data collection and analysis psychological processes (Blumenthal, To measure the startle reflex two Ag-AgCl 4 Cuthbert, Filion, Hackley, Lipp, & Van Boxtel, mm electrodes were placed on the obicularis 2005). oculi muscle below the left eyelid. The Test-retest reliability of the emotion ground electrode was placed at the AFz site. modulated startle response may not be Preparation of the sensors involved briskly stable over time; a study by Larson et al. brushing the skin with medical gauze, (2000) found low reliability at a second cleaning the area with rubbing alcohol, assessment. However, replication of this gently exfoliating with saline gel, and study that includes subjective ratings of brushing the skin once more with gauze. arousal and valence to stimuli are needed to EMG sensors were filled with Quickgel. help understand why the reliability of the Impedance levels of the electrodes below 20 emotion modulated startle appears to be kΩ were accepted (Larson et al., 2000). EMG low. A study by Lee et al. (2009) found low data was obtained with a Brain Products Inc, test-retest reliability of the emotion QuickAmp 72 System. Brain Vision Recorder regulated startle response, with decreased was used to record the data and Brain Vision sensitivity to regulation, compared to a Analyzer was used to process and score the corrugator measure. The startle paradigm startle responses. Data was sampled at 1000 measures emotion over a span of Hz. Signals were digitally filtered offline with milliseconds, which may account for its high and low pass frequency filters at 30 Hz variability across assessments, whereas and 500 Hz (48 dB/octave roll-off) with a 60 corrugator activity is measured over a larger Hz Notch filter. Startle EMG was rectified span of time and is therefore likely to be and smoothed with a 20 ms moving window more consistent. To be confident in the average. The data was baseline corrected, ability of the startle paradigm to measure subtracting the average value of the 50 ms emotion modulation and regulation over baseline period (before the probe onset) time more research needs to be done testing from all time points in the segment. The reliability, and the factors that may influence peak startle magnitude was determined it. within a window of time extending from the time of probe onset to 120 ms. Trials were excluded if the blink began prior to 15 ms following the probe, if there was excess

UBCUJP – In Press – Volume 1

CLINICAL Speed Research Report

0.3 subjective ratings of valence and arousal for

0.2 the images matched the normative ratings.

0.1

scores) - 0 Emotion Modulation A one-way repeated measures ANOVA was -0.1 conducted to examine the predicted -0.2 emotion modulation effects on the startle

BlinkMagnitude (z -0.3 response to probe A, which was presented 3 -0.4 s after picture onset, before the regulation Negative Neutral Positive Valence cue. The effect of valence was non- Figure 1. Differences in blink magnitude by affective significant, F(2,8) = 3.05, ns, startle blink picture type. Measured 3 s after picture onset. magnitudes for each picture type were not significantly different. The hypothesized

linear trend of negative>neutral>positive Results was examined using a linear trend contrast Image Ratings on valence and revealed a significant linear A one-way repeated measures ANOVA was trend, F(1,4) = 8.39, p< 0.05, blink conducted to evaluate the participant’s magnitudes did follow the expected pattern subjective ratings of picture valence. A of largest in magnitude to negative images significant effect for valence was revealed, and smallest in magnitude to positive F(2,10) = 52.28, p< 0.001. Follow-up paired t- images, supporting the argument that the tests revealed a significant difference of startle response is an aversive reflex. Follow- valence between negative and positive, t(5) up paired t-tests revealed a significant = -7.41, p = 0.001, neutral and negative, t(5) difference between negative and positive, = 8.48, p< 0.001, and neutral and positive, t(4) = 2.90, p< 0.05 while the difference t(5) = -5.42, p< 0.05. Therefore, participants between negative and neutral, t(4) = 0.83, ns rated each picture type as significantly and neutral and positive, t(4) = 1.48, ns were different from each other on emotional non-significant (see Figure 1). Therefore, valence. A one-way repeated measures blink magnitudes to negative pictures were ANOVA was conducted to evaluate the significantly larger than to positive pictures. participant’s subjective ratings of arousal. A However, no significant differences in blink significant effect for arousal was revealed, magnitudes were found between neutral F(2,10) = 34.37, p< 0.001. Follow-up paired t- and positive or neutral and negative images. tests revealed a significant difference for arousal between neutral and negative t(5) = Emotion Regulation -9.44, p< 0.001, and neutral-positive t(5) = - A 3 (image valence) x 2 (regulation 6.08, p< .005, while the difference between condition) repeated measures ANOVA was negative and positive, t(5) = 2.14, ns were conducted for the startle response to probe non-significant. In other words, participants B, presented at 7 s following picture onset. rated the positive and negative pictures as This revealed a significant main effect for high in emotional arousal, while the neutral valence, F(2,8) = 7.08, p< 0.05, blink images were rated as low in emotional magnitudes were significantly different arousal. Accordingly, the participant’s across picture types. A main effect for regulation approached significance, F(1,4) =

UBCUJP – In Press – Volume 1

CLINICAL PHYSIOLOGICAL EMOTION REGULATION Research Report

5.08, p = 0.087, meaning that differences in 0.8 Negative blink magnitude between the regulation Positive

0.6 conditions showed a trend, but were not Neutral

substantial enough to show significant scores) 0.4 - differences. The interaction between valence and regulation was non-significant, 0.2 F(2,8)=2.74, ns. Therefore, blink magnitudes 0

were not significantly affected by the BlinkMagnitude (z relationship between picture type and -0.2 regulation cue. The linear contrast for -0.4 valence was significant with Maintain Suppress negative>positive>neutral, F(1,4) = -8.69, p< Regulation Cue 0.05. Concretely, a consistent linear trend Figure 2. Changes in blink magnitude relative to the was found for blink magnitudes and picture affective picture type and regulation cue condition. Measured 7 s after picture onset. type at both probe times, before and after the regulation cue, highlighting the stability successfully, and that the pictures were of this effect. Paired t-tests were conducted eliciting the expected response, we expected to test the hypothesized regulation effects to replicate the findings of other emotion for the negative and positive images (see modulation research (Bradley et al., 1991; Figure 2). A significant difference was found Sanchez-Navarro, Martinez-Selva, Torrente, between negative maintain and negative & Roman, 2008; Vrana et al., 1988). The decrease, t(4) = 5.01, p< 0.05, with blink startle magnitude was significantly larger for magnitudes being significantly smaller in the negative compared to positive images. decrease condition than the maintain However, responses to the neutral images condition. The difference between positive were not significantly different from either maintain and positive decrease were non- the positive or negative images. Failing to significant, t(4) = 1.06, ns. Blink magnitudes find differences in response magnitude for were not significantly different in size neutral and positive images has been a between the decrease and maintain problem in other emotion regulation studies conditions. As expected there was no (Dillon & La Bar, 2005; Driscoll et al., 2008). significant difference in blink magnitudes This finding may indicate that the positive between neutral maintain and neutral images were not interpreted as positively as suppress conditions. Therefore, significant expected by the participants. differences in blink magnitude were found The lack of significant differences only during regulation to the negative between negative pictures to neutral images. pictures could be due to a lack of

habituation to the startle probe at the Discussion beginning of the paradigm. The first probed The initial emotional response to the images in both trials were neutral, and pictures, before any regulation cues were responses to these images were typically given, as measured by probe A, was an much larger than for the rest of the neutral important characteristic of this study. To pictures in the task. Therefore, a few large help assess that the paradigm was working startle responses in the neutral condition

UBCUJP – In Press – Volume 1

CLINICAL Speed Research Report

could be responsible for driving the mean Current research has not found the up. Leaving out the first trial in the paradigm expected emotion regulated startle response has been used in previous research as it has to positive images (Dillon & La Bar, 2005; been found to be larger than all other Driscoll et al., 2008). Jackson et al. (2000) responses, and may be useful in future has proposed that this may be due to a research to avoid habituation effects (Larson difficulty in eliciting positive emotion in a et al., 2000). Another possible reason laboratory setting. Genuine positive emotion significant differences were not found for may be harder to elicit in participants than positive and negative images to neutral is negative emotion, as the participants are because of the small sample size in this sitting in a dark room, unaware of what they study. Due to the small sample size, the will see next, with loud noises occasionally power to detect smaller significant startling them. Also, to ensure that arousal is differences was low. matched in both negative and positive Emotion regulation to the picture images, erotic images are often used stimuli was also measured by the startle because other positive images are often response. The first hypothesis stated that in rated low on arousal. Although the startle the suppress negative condition the startle response is supposed to be smallest when response would be attenuated compared to viewing erotic images (Lang et al., 1990), it is the maintain negative condition. This possible that individuals may vary on their hypothesis was supported, during the responses to the erotic images in the lab suppress negative condition the startle setting, depending on their comfort level. response was significantly smaller than Another explanation for the failure to during the maintain negative condition. The find the expected regulation to positive startle blink paradigm successfully measured images comes from Dillon and La Bar (2005) voluntary changes in emotional response to who suggested that the emotion regulated a regulation cue for the negative images. startle response is arousal-dependent. The second hypothesis stated that for the Therefore, stimuli that are arousing, positive images, startle blink magnitude regardless of valence, should show an would be potentiated during the suppress attenuated startle response when positive condition compared to the maintain suppressed compared to maintained, in positive condition. As others have found, opposition to the aversive matching this hypothesis was not confirmed; the hypothesis. However, both the positive and startle response was smaller during the negative stimuli in the task were matched on suppress condition than the maintain arousal for normative and subjective ratings, condition for positive images, however, this but positive images failed to show any difference was not significant (Dillon & La significant differences between the suppress Bar, 2005; Driscoll et al., 2008). In this study and maintain conditions, while the negative the startle blink paradigm was not a images did show significant differences sensitive measure of emotion regulation to between condition. Therefore, arousal does positive affective stimuli. No significant not appear to fully explain why positive differences were expected for startle blink emotion regulation has not followed the magnitudes between the suppress neutral pattern expected by the biphasic (appetitive- and the maintain neutral conditions, and aversive) theory of emotion. indeed no differences were found.

UBCUJP – In Press – Volume 1

CLINICAL PHYSIOLOGICAL EMOTION REGULATION Research Report

An important limitation of this study, and drinking, gambling, and self-injury and others, that have attempted to measure (Klonsky, 2008). As most research regarding positive and negative emotion regulated emotion regulation is self-report it is critical startle is the small size of the sample (Dillon to address if these emotion regulation & La Bar, 2005; Driscoll et al., 2008). The deficits are also present at the physiological magnitude of the emotion modulated and level, in order to better develop and assess regulated startle varies widely across treatment strategies that improve emotion individuals (Lang et al., 1990), therefore, a regulation cognitively, behaviourally, and larger sample size is needed before making physically. However, due to inconsistencies any strong conclusions about the results. in the research and insufficient data to The small sample size reported in this and support the expected hypotheses, more other research may play a large role in the research needs to be done using the startle failure to gather significant or consistent paradigm on positive and negative emotion findings of emotion regulation to positive regulation, with large sample sizes, to stimuli. properly understand the mechanisms at This study only used female work that influence the startle response undergraduate subjects, as have others before it can be used to provide reliable (Eippert et al., 2007; Ray et al., 2010), which information about clinical samples. limits the generalizability of the results. Future studies on the emotion Females were used exclusively in this study regulated startle response should utilize a as they were participants in part of a larger larger sample size, balanced paradigms for study involving behaviours most common to picture valence, and matched subjective and women. Although previous research that has normative arousal for positive and negative included both sexes has found no significant stimuli in order to successfully extend upon differences of gender (Jackson et al., 2000) it current knowledge on the nature of the would be beneficial to include both sexes in startle response to emotion regulation. It a study of the emotion regulated startle may also be beneficial to look for more response with positive and negative stimuli, effective ways to elicit positive emotion in to be able to further generalize the findings the lab setting, beyond the IAPS picture set. and to test for any possible differences in Moreover, fear conditioning has been shown emotion regulation between gender. How to modulate the startle response in the lab these results relate to other populations, for (Greenwald et al., 2007; Lissek et al., 2007); example, non-educated individuals, older perhaps associating a stimulus with a adults or clinical populations is unclear. It positive experience or reward could also be would be beneficial if these findings could be used to modulate the startle response, replicated with more varied samples. eliciting a more genuine positive emotion in In the introduction it was suggested the lab. In conclusion, the startle blink that the emotion regulated startle response paradigm is a promising physiological could potentially be a useful measure of measure of emotion regulation, however emotion regulation abilities/deficits in more research needs to be conducted with clinical samples. Emotion dysregulation, or larger samples before it can be used to the inability to regulate one’s emotions in a assess dysregulation in clinical populations. healthy or adaptive way has been linked to destructive behaviours such as binge eating

UBCUJP – In Press – Volume 1

CLINICAL Speed Research Report

Declaration of Conflicting Interests threat-related stimuli. Human Brain Mapping, The author declared they have no conflicts 28, 409-423. Funayama, E. S., Grillon, C., Davis, M., & Phelps, E. of interests with respect to their authorship A. (2001). Double dissociation in the affective or the publication of this article. modulation of startle in humans: Effects of unilateral temporal lobectomy. Journal of Cognitive Neuroscience, 13(6), 721-729. References Gross, J. J. (1998). The emerging field of emotion Aldao, A., & Nolen-Hoeksema, S. (2010). Specificity regulation: An integrative review. Review of of cognitive emotion regulation strategies: A General Psychology, 2(3), 271-299. transdiagnostic examination. Behaviour Greenwald, M. K., Bradley, M. M., Cuthbert, B. N., & Research and Therapy, 48(10), 974-983. doi: Lang, P. J. (2007). Startle potentiation: Shock 10.1016/j.brat.2010.06.002 sensitization, aversive learning, and affective Blumenthal, T. D., Cuthbert, B. N., Filion, D. L., picture modulation. Behavioural Neuroscience, Hackley, S., Lipp, O. V., & Van Boxtel, A. 112(5), 1069-1079. doi: 10.1037/0735- (2005). Committee report: Guidelines for 7044.112.5.1069 human startle eyeblink electromyogtaphic Harrison, A., Sullivan, S., Tchanturia, K., & Treasure, studies. Psychophysiology, 42(1), 1-15. doi: J. (2010). Emotional functioning in eating 10.1111/j.1469-8986.2005.00271.x disorders: Attentional bias, emotion recognition Bradley, M. M., Cuthbert, B. N., & Lang, P. J. (1991). and emotion regulation. Psychological Startle and emotion: Lateral acoustic Medicine, 40, 1887-1897. probes and the bilateral blink. Jackson, D. C., Malmstadt, J. R., Larson, C. L., & Psychophysiology, 28(3), 285-295. doi: Davidson, R. J. (2000). Suppression and 10.1111/j.1469-8986.1991.tb02196.x enhancement of emotional responses to Bradley, M. M., & Lang, P. J. (1994). Measuring unpleasant pictures. Psychophysiology, 37, emotion: The self-assessment manikin and the 515-522. doi:10.1017/S0048577200990401 semantic differential. Journal of Therapy and Kim, S. J., & Hamann, S. (2007). Neural correlates of Experimental Psychiatry, 25, 49-59. positive and negative emotion regulation. Clyne, C., Latner, J. D., Gleaves, D. H., & Blampied, Journal of Cognitive Neuroscience, 19(5), 776- N. M. (2010). Treatment of emotional 798. doi: 10.1162/jocn.2007.19.5.776 dysregulation in full syndrome and Klonsky, E. D. (2008). The functions of deliberate subthreshold binge eating disorder. Eating self-injury. A review of the evidence. Clinical Disorders, 18, 408-242. Psychology Review, 27, 226-239. Cook, E. W., Hawk, L. W., Davis, T. L., & Stevenson, Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1990). V. E. (1991). Affective individual differences Emotion, attention, and the startle reflex. and startle reflex modulation. Journal of Psychological Review, 97(3), 377-395. doi: Abnormal Psychology, 100, 5-13. 10.1037/0033-295X.97.3.377 Davidson, R. J. (2000). Affective style, Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1999). psychopathology, and resilience: Brain International affective picture system mechanisms and plasticity. American (IAPS): Technical manual and affective Psychologist, 55(11), 1196-1214. doi: ratings. NIMH Centre for the Study of Emotion 10.1037/0003-066X.55.11.1196 and Attention. Driscoll, D., Tranel, D., & Anderson, S. W. (2008). Larson, C. L., Ruffalo, D., Nietert, J. Y., & Davidson, The effects of voluntary regulation of positive R. J. (2005). Stability of emotion-modulated and negative emotion on psychophysiological startle during short and long picture responsiveness. International Journal of presentation. Psychophysiology, 42, 604-610. Psychophysiology, doi: doi: 10.1111/j.1469-8986.2005.00345.x 10.1016/j.ijpsycho.2008.03.012. Larson, C. L., Ruffalo, D., Nietert, J. Y., & Davidson, Dillon, D. G., & La Bar, K. S. (2005). Startle R. J. (2000). Temporal stability of the emotion- modulation during conscious emotion modulated startle response. Psychophysiology, regulation is arousal-dependent. Behavioural 37, 92-101. doi: 10.1017/S0048577200981344 Neuroscience, 119(4), 1118-1124. Lee, H., Shackman, A. J., Jackson, D. C., & Davidson, Eippert, F., Veit, R., Weiskopf, N., Erb, M., R. J. (2009). Test-retest reliability of voluntary Birbaumer, N., & Anders, S. (2007). emotion regulation. Psychophysiology, 46, 874- Regulation of emotional responses elicited by 879.

UBCUJP – In Press – Volume 1

CLINICAL PHYSIOLOGICAL EMOTION REGULATION Research Report

Lissek, S., Orme, K., Mcdowell, D. J., Johnson, L. L., Sanchez-Navarro, J. P., Martinez-Selva, J. M., Luckenbaugh, D. A., Baas, J. M., Cornwell, B. Torrente, G., & Roman, F. (2008). R., & Grillon, C. (2007). Emotion regulation of Psychophysiological, behavioral, and cognitive potentiated startle across affective picture and indices of the emotional response: A factor- threat-of-shock paradigms. Biological analytic study. The Spanish Journal of Psychology, 76, 124-133. Psychology, 11(1), 6-25. Miltner, W., Matjak, M., Braun, C., Diekmann, H., & Vrana, S. R., Spence, E. L., & Lang, P. J. (1988). The Brody, S. (1994). Emotional qualities of odors Startle Probe Response: A New Measure of and their influence on the startle reflex in Emotion? Journal of Abnormal Psychology, humans. Psychophysiology, 31, 107-110. 97(4), 487-491. Ray, R. D., McRae, K., Ochsner, K. N., & Gross, J. J. Whiteside, U., Chen, E., Neighbors, C., Hunter, D., (2010). Cognitive reappraisal of negative Lo, T., & Larimer, M. (2007). Difficulties : Converging evidence from EMG and regulating emotions: Do binge eaters have self-report. Emotion, 10(4), 587-592. doi: fewer strategies to modulate and tolerate 10.1037/a0019015 negative affect? Eating Behaviours, 8, 162- 169.

UBCUJP – In Press – Volume 1