- Email: [email protected]
Facial EMG as an indicator of palatability in humans
Physiology & Behavior 68 (1999) 31–35
Facial EMG as an indicator of palatability in humans Senqi Hu,* Kathryn A. Player, Kathleen A. Mcchesney, Maria D. Dalistan, Catherine A. Tyner, and Jason E. Scozzafava Department of Psychology, Humboldt State University, Arcata, CA 95521, USA Received 25 June 1998; received in revised form 27 May 1999; accepted 8 July 1999
Abstract Two experiments were conducted to observe the sensory hedonic responses and facial EMG activities elicited from different tastes. In Experiment 1, 25 subjects tasted flavors of apple juice, Gatorade, water, soybean milk, and pickle juice. In Experiment 2, 21 subjects tasted a sugar solution, a salt solution, and water. Subjects’ sensory hedonic reports and facial EMG activity in the levator labii superioris/ alaeque nasi region were recorded in each experiment. The results from Experiment 1 showed that subjects reported significantly higher ratings of palatability to the apple juice, Gatorade, and water than to the soybean milk and pickle juice. The results from Experiment 2 indicated that the subjects reported significantly higher ratings of palatability to the water and sugar solution than to the salt solution. EMG recordings in both experiments showed that the negative hedonic sensations were associated with higher EMG activity in the levator labii muscle region, and that the positive hedonic sensations were associated with lower EMG activity in the same muscle region. It is concluded that facial EMG activity can be used as an indicator of palatability in humans. © 1999 Elsevier Science Inc. All rights reserved. Keywords: Electromyogram; Taste; Flavor; Hedonics; Palatability
1. Introduction In previous taste hedonic studies, researchers have used both self-reports of palatability and observations of facial expressions to assess human reactivity to a flavor or odor [1]. However, the observational technique for evaluating taste reactivity involves observing numerous patterns of facial expression and exhaustive schemes for detecting patterns of facial motion. For example, there are 25 detectable facial motion features that need to be evaluated in terms of frequency and duration of occurrence [1]. Also, subjective assessment of facial motion features is likely to generate different results from different evaluators. Furthermore, it is questionable whether or not facial motion is sensitive enough to reflect covert reactivities of facial muscles [2]. Electromyographic (EMG) recordings may provide alternative means for assessing facial expressions associated with taste hedonics. EMG recordings are unobtrusive and capable of detecting changes in emotional processes that are too subtle to detect for observational techniques [2–4]. Also, errors that may arise from an observer’s subjective assessment are eliminated when EMG signals are recorded objectively.
* Corresponding author. Tel.: 707-826-5262; Fax: 707-826-4993. E-mail address: [email protected]
Considerable research has demonstrated that surface EMG is useful in discriminating among different types of emotions. It was reported that EMG activity in the zygomatic muscle, which elevates the cheeks to form a smile, increased when individuals imagined pleasant thoughts, and that EMG activity in the corrugator muscle, which furrows the brows, increased during the imagination of unpleasant thoughts [5–7]. It was also found that EMG activity in the corrugator region increased when female subjects were viewing slides of unpleasant social scenes, and decreased when viewing pleasant ones [8]. Recent studies have shown that disgust imagery was associated with increased EMG activity in the levator labii superioris/alaeque nasi muscle, which lifts the middle of the upper lip and wrinkles the nose [2,9]. The present study observed sensory hedonic responses to different tastes and associated facial EMG activities. Two experiments were conducted. In Experiment 1, 25 subjects tasted flavors of apple juice, Gatorade, water, soybean milk, and pickle juice. In Experiment 2, 21 subjects tasted flavors of sugar solution, salt solution, and water. Subjects’ facial EMG activity and subjective hedonic reports were recorded in both experiments. Because previous findings indicated that disgust is associated with increased EMG activity in the levator labii muscle [2,9], we hypothesized that subjects’ EMG activity in the levator labii muscles would significantly increase when tasting the flavors with high ratings of negative sensory hedonics.
0031-9384/99/$ – see front matter © 1999 Elsevier Science Inc. All rights reserved. PII: S0031-9384(99)00 1 4 3 - 2
32
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
2. Experiment 1 2.1 Method 2.1.1. Subjects Twenty-five undergraduate students (9 men and 16 women) at Humboldt State University participated in the experiment. Subjects were recruited from various psychology courses and received extra credit as compensation. The subjects’ ages ranged from 18 to 30 years (mean 5 20.4 years, SD 5 2.7). None of the subjects reported having any allergies to certain tastes. The experimental procedures were approved by the Committee for the Protection of Human Subjects of Humboldt State University, and informed written consent was obtained from each subject. 2.1.2. Materials and procedures The five tastes used in the experiment were: apple juice (Tree Top Apple Juice, Tree Top Inc., Selah, WA), Gatorade (Lemon-Lime Gatorade, The Gatorade Company, Chicago, IL), water (Alhambra Mountain Spring Water, McKesson Water Products Co., Pasadena, CA), pickle juice (Del Monte Slow Cured Whole Pickle Juice, Del Monte Foods, San Francisco, CA), and soybean milk (Natural Unflavored Tofu Shop Soy Milk Specialty Foods Inc., Arcata, CA). Five plastic graduated syringes were used to measure the amount of liquid. The experiment was conducted in a private suite of a laboratory in the Department of Psychology. All 25 subjects had one laboratory session, in which they sat in a chair with a head rest. Two 6-mm silver–silver chloride cutaneous disk electrodes (Grass Instrument Division, Astro-Med, Inc., West Warwick, RI) were placed on the subjects’ left levator labii (superioris/alaeque nasi muscle) region to record facial EMG activity. Figure 1 illustrates the locations of two electrode placement. The levator labii superioris/alaeque nasi is
FIG. 1. Placement of electrodes on levator labii superioris alaeque nasi muscle region for surface EMG recordings.
a thin triangular muscle located by the side of the nose (Fig. 1). It begins by the inner margin of the orbit, and ends at upper lip and cartilage of the ala of the nose [10]. The electrodes were placed parallel to the muscle fibers with a distance of 6 mm between them. A third ground electrode was placed on the mastoid behind the left ear. Recording sites were prepared with rubbing alcohol and a mild skin abrasive. The electrodes were connected to an electromyogram amplifier model (EMG100B, Biopac Systems, Inc., Goleta, CA) with a frequency passband of 100–500 Hz. Then the EMG signals from the amplifier were digitized at a 2048 Hz sampling rate via a 16-bit analog/digital converter (UIM100, Biopac Systems, Inc., Goleta, CA) using a data acquisition and process software (MP100 Data Acquisition System, Biopac Systems, Inc., Goleta, CA) on an IBM-compatible Pentium computer. Each subject first sat still in the chair while recording 64 s of baseline EMG signals. Then, one of the experimenters infused 15 mL of the first liquid flavor into the subject’s mouth with a plastic syringe. The subject held the liquid in their mouth for 64 s while their facial EMG activities were recorded. Immediately after the liquid holding period, the subject read and made a single mark on a 100-mm horizontal visual-analog scale of palatability to the flavor. The scale ranged from 0 to 100 with “extremely like” as an anchor equal to 100 located at the right end of the scale, and “extremely dislike” as an anchor equal to 0 located at the left end of the scale. Then the subject spat out the liquid into a basin and rinsed their mouth thoroughly with water. Afterward, the subject rested for a period of 5 min. The subject repeated the same cycle of recording a baseline of EMG activity, recording a tasting period of EMG activity, marking the visual scale, and resting for 5 min after each of the remaining flavors. The five flavors tasted by each subject were: apple juice, Gatorade, water, soybean milk, and pickle juice. The order of the flavors was counterbalanced with a 5 3 5 Latin square design. 2.1.3. Data analysis The scores of palatability ratings on the visual scale were determined by measuring the length in millimeters from the left end of the scale to the mark. The mean scores for each flavor were then statistically compared. The EMGs were analyzed using the Fast Fourier Transform (FFT) spectral analysis technique (MP100 Data Acquisition system, Biopac Systems, Inc., Goleta, CA). Each 64 s of EMG data that was originally collected at the sampling rate of 2048 Hz during the baseline or flavor tasting periods delineated a time series epoch. After FFT analysis, the time series epoch was transformed into a frequency epoch with the range of 1 to 1024 Hz. Then, the numerical values were squared to obtain the spectral power, i.e., spectral density in mv2 for each frequency unit of the frequency epoch. The mean values of spectral power of the epoch were then obtained for the frequency band of 100 to 250 Hz. Because the numerical values were squared during the process
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
33
of obtaining spectral power, these values generated large variations between subjects. The ratio of EMG frequency band of 100 to 250 Hz between tasting and baseline periods was then calculated in the purpose of reducing variations in statistical analysis. 2.2. Results Table 1 presents the means and standard deviations for ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for the five different tastes (Table 1). As can be seen in Table 1, subjects reported the highest ratings of palatability to the apple juice, followed by Gatorade, water, soybean milk, and pickle juice. The subjects generated the lowest ratios of EMG spectral power between tasting and baseline periods of water, followed by apple juice, Gatorade, soybean milk, and pickle juice. The scatter plot of ratios of EMG spectral power against ratings of palatability when all data from five tasting conditions were pooled (Fig. 2) indicate a negative relationship (r 5 20.53, p , 0.0001). A univariate repeated measures analysis of variance on ratings of palatability indicated that there were significant differences among the tasting conditions, F(4, 96) 5 18.50, p , 0.001). Further univariate follow-up comparisons (paired t-tests) showed that the subjects during the apple juice tasting condition reported significantly higher ratings of palatability than during the soybean milk tasting condition, t(24) 5 4.65, p , 0.001, and than during the pickle juice tasting condition, t(24) 5 6.33, p , 0.001. The subjects during the Gatorade tasting condition reported significantly higher ratings of palatability than during the soybean milk tasting conditions, t(24) 5 2.72, p , 0.05, and than during the pickle juice tasting condition, t(24) 5 5.21, p , 0.001. The subjects during the water tasting condition reported significantly higher ratings of palatability than during the soybean milk tasting condition, t(24) 5 2.88, p , 0.01, and than during the pickle juice tasting condition, t(24) 5 5.63, p , 0.001. A univariate repeated-measures analysis of variance on ratios of facial EMG spectral power between tasting and baseline periods indicated that there were significant differences among the tasting conditions, F(4, 96) 5 8.79, p ,
Table 1 Means and standard deviations of ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for five different tastes Ratios of EMG spectral power
Ratings of palatability Flavors
n
Mean
SD
Mean
SD
Apple juice Gatorade Water Soybean milk Pickle juice
25 25 25 25 25
78.44 62.00 61.40 42.68 28.68
23.10 22.30 18.71 28.23 28.77
1.34 1.48 1.33 1.82 2.45
0.38 0.42 0.33 0.98 1.29
FIG. 2. Scatter plot of ratios of EMG spectral power versus ratings of palatability to the five flavors of 25 subjects.
0.001. Further univariate follow-up comparisons (paired ttests) showed that the subjects during the apple juice tasting condition generated significantly lower ratios of EMG spectral power between tasting and baseline periods than during the soybean milk tasting condition, t(24) 5 2.60, p , 0.05, and than during the pickle juice tasting condition, t(24) 5 4.14, p , 0.001. The subjects during the water tasting condition generated significantly lower ratios of facial EMG spectral power between tasting and baseline periods than during the soybean milk tasting condition, t(24) 5 2.30, p , 0.05, and than during the pickle juice tasting condition, t(24) 5 3.96, p , 0.001. 3. Experiment 2 3.1 Method 3.1.1 Subjects The same criteria and procedures in Experiment 1 were used in Experiment 2 to recruit 21 undergraduate students (6 men and 15 women). Their ages ranged from 18 to 30 years (mean 5 20.8 years, SD 5 1.9). 3.1.2. Materials and procedures The same equipment and procedures in Experiment 1 were used in Experiment 2. However, there were three tastes used in Experiment 2 that were: water (Alhambra Mountain Spring Water, McKesson Water Products Co., Pasadena, CA), a sugar solution: 20 g of sugar (Spreckels All Natural Sugar, Imperial Holly Corporation, Sugar Land, TX) dissolved in 150 mL water, and a salt solution: 5 g of salt (Morton Salt, Morton International Inc., Chicago, IL) dissolved in 150 mL water. The order of the tastes was counterbalanced with a 3 3 3 Latin square design.
34
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
3.2. Results Table 2 presents the means and standard deviations for ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for the three different tastes (Table 2). As can be seen in Table 2, subjects reported the highest ratings of palatability to water, followed by sugar and salt. However, the subjects generated the lowest ratios of EMG spectral power between tasting and baseline periods to water, followed by sugar and salt. The scatter plot of ratios of EMG spectral power against ratings of palatability when all data from three tasting conditions were pooled (Fig. 3) indicate a negative relationship (r 5 20.49, p , 0.001). A univariate repeated-measures analysis of variance on ratings of palatability indicated that there were significant differences among the tasting conditions, F(2, 40) 5 46.99, p , 0.001. Further univariate follow-up comparisons (paired t-tests) showed that subjects during the salt-tasting condition reported significantly lower ratings of palatability than during the water-tasting condition, t(20) 5 9.21, p , 0.001, and than during the sugar-tasting condition, t(20) 5 7.96, p , 0.001. No significant difference was found for ratings of palatability between the sugar and water tasting conditions. A univariate repeated-measures analysis of variance on ratios of facial EMG spectral power between tasting and baseline periods indicated that there were significant differences among the tasting conditions, F(2, 40) 5 10.70, p , 0.001. Further univariate follow-up comparisons (paired t-tests) showed that the subjects during the salt tasting condition generated significantly higher ratios of facial EMG spectral power between tasting and baseline periods than during the water tasting condition, t(20) 5 3.98, p , 0.001, and than during the sugar-tasting condition, t(20) 5 3.21, p , 0.01. No significant difference was found on ratios of facial EMG spectral power between tasting and baseline periods between water- and sugar-tasting conditions. 4. General discussion The present study investigated the sensory hedonic responses and associated facial EMG activities to different tastes. The results from the two experiments showed that
Table 2 Means and standard deviations of ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for three different tastes Ratios of EMG spectral power
Ratings of palatability Flavors
n
Mean
SD
Mean
SD
Salt Sugar Water
21 21 21
22.57 67.00 72.76
21.81 18.56 24.17
1.87 1.31 1.25
0.79 0.58 0.41
FIG. 3. Scatter plot of ratios of EMG spectral power versus ratings of palatability to the water, sugar, and salt solutions of 21 subjects.
subjects’ ratings of palatability on the 0 to 100 hedonic scale ranged form 61.4 to 78.44 points while tasting apple juice, Gatorade, water, and sugar solution; and ranged from 22.57 to 42.68 points while tasting pickle juice, soybean milk, and salt solution. Because 50 points on the scale indicated a neutral hedonic level, these results suggest that apple juice, Gatorade, water, and sugar solution seem to generate positive hedonic responses, while pickle juice, soybean milk, and salt solution tend to generate negative hedonic responses. The experimental results also revealed that subjects generated higher ratios of facial EMG spectral power in the levator labii region between tasting and baseline periods to flavors with negative hedonic levels (such as pickle juice, soybean milk, and salt solution) than to flavors with positive hedonic levels (such as apple juice, Gatorade, water, and sugar solution). Thus, the negative hedonic sensations were associated with higher levator labii EMG activity, while the positive hedonic sensations were associated with lower levator labii EMG activity. These results suggest that EMG activity in facial levator labii muscles can be used as an indicator of palatability. The results of the present study are not only consistent with those in a recent study that showed disgust imagery is accompanied by increased EMG activities in the levator labii muscles [2], but may also support prior theoretical proposals that this particular muscle group has evolved as a functional necessity to protect the body from contamination, poisoning, or illness due to oral consumption [2,9,11,12]. It is known that the contraction of the levator labii muscle raises the upper lip and wrinkles the nose [2,8]. The contraction of the levator labii muscles occurs typically when it is necessary to avoid unpleasant odors by closing the nostrils and to expel noxious contents in the mouth by opening
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
the mouth. Previous research results from infant studies have also provided such theoretical support by noting that facial expressions of disgust existed in 2-day-old newborns when tasting solutions of citric acid, and that more positive facial expressions were observed when tasting solutions of sucrose [11]. The acts of using this muscle group surrounding the nose to detect or reject odoriferous foods and to open the mouth to dribble out unpalatable foods have been observed in both animal and human studies of taste reactivity and taste aversion [11,13]. However, it is important to point out that because present study recorded facial EMG activities from only one facial site, the levator labii muscle region, we are unable to say that the increase of EMG activity in the levator labii muscle region is the only indicator of palatability. To define the whole picture of facial EMG pattern associated with hedonic response, researchers need to record facial EMG signals from multiple sites in future sensory hedonic studies. Because previous studies have repeatedly demonstrated that laboratoryinduced unpleasant feelings are associated with high facial EMG activities in the corrugator regions [3,5,7,8], it is important to compare the differences between the EMG activities in the levator labii region and those in the corrugator region while tasting unpleasant flavors. Furthermore, the purpose of the present study was to investigate the relationship of palatability and facial EMG activities in the region of levator labii muscles. We were not attempting to define the facial EMG pattern of disgust. Because disgust is defined as a strong feeling of dislike or finding a thing very unpleasant [14], we consider that the negative hedonic sensation is related to the emotional state of disgust. However, it is too early to conclude that the increased EMG activity in the levator labii muscle region reflects only the emotional state of disgust. The eventual understanding of how disgust is manifested as an identified muscular pattern relies on the analysis of multiple facial EMG recordings from different facial regions. It is also impossible to use the data of the present study to verify other negative emotion states such as anger or sadness. We believe that negative hedonic experience is like an emotional experience that contains expressive, experiential, and physiological components [15–17]. These components are interrelated when emotional reactions are evoked [16– 19]. With the EMG recording, a physiological measure, researchers can obtain quantitative data, avoid human errors, and save time in coding facial motion patterns. Therefore, it
35
is recommended that EMG recording methods be used in future taste hedonic studies. References [1] Steiner JE, Lidar-Lifschitz D, Perl E. Taste and odor: reactivity in depressive disorders, a multidisciplinary approach. Percept Mot Skills 1993;77:331–1346. [2] Vrana SR. The psychophysiology of disgust; differentiating negative emotional context with facial EMG. Psychophysiology 1993;30:279– 86. [3] Cacioppo JT, Martzke JS, Petty RE, Tassinary LG. Specific forms of facial EMG response index emotions during an interview: from Darwin to the continuous flow hypothesis of affect-laden information processing. J Pers Soc Psychol 1988;54:592–604. [4] Dimberg U. Facial electromyography and emotional reactions. Psychophysiology 1990;27:481–94. [5] Schwartz GE, Ahern L, Brown SL. Lateralized facial muscle response to positive and negative emotional stimuli. Psychophysiology 1979;16:561–71. [6] Schwartz GE, Brown SL, Ahern L. Facial muscle patterning and subjective experience during affective imagery: Sex differences. Psychophysiology 1980;17:75–82. [7] Schwartz GE, Fair PL, Salt P, Mandel MR, Klerman GL. Facial muscle patterning to affective imagery in depressed and nondepressed subjects. Science 1976;192:489–91. [8] Cacioppo JT, Bush LK, Tassinary LG. Microexpressive facial actions as a function of affective stimuli: replication and extension. Pers Soc Psychol Bull 1990;18:515–26. [9] Vrana SR. Startle reflex response during sensory modality specific disgust, anger, and neutral imagery. J Psychophysiol 1994;8:211–18. [10] Gray H, Pick TP, Howden R. Gray’s Anatomy. Philadelphia, PA: Running Press, 1974. [11] Fox NA, Davidson RJ. Taste-elicited changes in facial signs of emotion and asymmetry of brain electrical activity in human newborns. Neuropsychology 1986;24:417–22. [12] Rozin P, Fallon AE. A perspective on disgust. Psychol Rev 1987;94: 23–41. [13] Parker LA. Rewarding drugs produce taste avoidance, but not taste aversion. Neurosci Biobehav Rev 1995;19:143–50. [14] Oxford American Dictionary. New York: Oxford University Press, 1980. [15] Izard CE, Kagan J, Zajonc RB. Emotions, Cognition, and Behavior. Cambridge, MA: Cambridge University Press, 1984. [16] Lang PJ. A bio-informational theory of emotional imagery. Psychophysiology 1979;16:495–512. [17] Zajonc RB. Feeling and thinking: preferences need no inferences. Am Psychol 1980;35:151–75. [18] Ekman P. Cross-cultural studies of facial expressions. In: Ekam, P, editor. Darwin and Facial Expression. New York: Academic Press, 1973, pp. 169–222. [19] Lazarus RS. Thoughts on the relations between emotion and cognition. In: Scherer KL, Ekman P., editors. Approaches to Emotion. Hillsdale, NJ: Erlbaum, 1984, pp. 247–57.
Facial EMG as an indicator of palatability in humans Senqi Hu,* Kathryn A. Player, Kathleen A. Mcchesney, Maria D. Dalistan, Catherine A. Tyner, and Jason E. Scozzafava Department of Psychology, Humboldt State University, Arcata, CA 95521, USA Received 25 June 1998; received in revised form 27 May 1999; accepted 8 July 1999
Abstract Two experiments were conducted to observe the sensory hedonic responses and facial EMG activities elicited from different tastes. In Experiment 1, 25 subjects tasted flavors of apple juice, Gatorade, water, soybean milk, and pickle juice. In Experiment 2, 21 subjects tasted a sugar solution, a salt solution, and water. Subjects’ sensory hedonic reports and facial EMG activity in the levator labii superioris/ alaeque nasi region were recorded in each experiment. The results from Experiment 1 showed that subjects reported significantly higher ratings of palatability to the apple juice, Gatorade, and water than to the soybean milk and pickle juice. The results from Experiment 2 indicated that the subjects reported significantly higher ratings of palatability to the water and sugar solution than to the salt solution. EMG recordings in both experiments showed that the negative hedonic sensations were associated with higher EMG activity in the levator labii muscle region, and that the positive hedonic sensations were associated with lower EMG activity in the same muscle region. It is concluded that facial EMG activity can be used as an indicator of palatability in humans. © 1999 Elsevier Science Inc. All rights reserved. Keywords: Electromyogram; Taste; Flavor; Hedonics; Palatability
1. Introduction In previous taste hedonic studies, researchers have used both self-reports of palatability and observations of facial expressions to assess human reactivity to a flavor or odor [1]. However, the observational technique for evaluating taste reactivity involves observing numerous patterns of facial expression and exhaustive schemes for detecting patterns of facial motion. For example, there are 25 detectable facial motion features that need to be evaluated in terms of frequency and duration of occurrence [1]. Also, subjective assessment of facial motion features is likely to generate different results from different evaluators. Furthermore, it is questionable whether or not facial motion is sensitive enough to reflect covert reactivities of facial muscles [2]. Electromyographic (EMG) recordings may provide alternative means for assessing facial expressions associated with taste hedonics. EMG recordings are unobtrusive and capable of detecting changes in emotional processes that are too subtle to detect for observational techniques [2–4]. Also, errors that may arise from an observer’s subjective assessment are eliminated when EMG signals are recorded objectively.
* Corresponding author. Tel.: 707-826-5262; Fax: 707-826-4993. E-mail address: [email protected]
Considerable research has demonstrated that surface EMG is useful in discriminating among different types of emotions. It was reported that EMG activity in the zygomatic muscle, which elevates the cheeks to form a smile, increased when individuals imagined pleasant thoughts, and that EMG activity in the corrugator muscle, which furrows the brows, increased during the imagination of unpleasant thoughts [5–7]. It was also found that EMG activity in the corrugator region increased when female subjects were viewing slides of unpleasant social scenes, and decreased when viewing pleasant ones [8]. Recent studies have shown that disgust imagery was associated with increased EMG activity in the levator labii superioris/alaeque nasi muscle, which lifts the middle of the upper lip and wrinkles the nose [2,9]. The present study observed sensory hedonic responses to different tastes and associated facial EMG activities. Two experiments were conducted. In Experiment 1, 25 subjects tasted flavors of apple juice, Gatorade, water, soybean milk, and pickle juice. In Experiment 2, 21 subjects tasted flavors of sugar solution, salt solution, and water. Subjects’ facial EMG activity and subjective hedonic reports were recorded in both experiments. Because previous findings indicated that disgust is associated with increased EMG activity in the levator labii muscle [2,9], we hypothesized that subjects’ EMG activity in the levator labii muscles would significantly increase when tasting the flavors with high ratings of negative sensory hedonics.
0031-9384/99/$ – see front matter © 1999 Elsevier Science Inc. All rights reserved. PII: S0031-9384(99)00 1 4 3 - 2
32
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
2. Experiment 1 2.1 Method 2.1.1. Subjects Twenty-five undergraduate students (9 men and 16 women) at Humboldt State University participated in the experiment. Subjects were recruited from various psychology courses and received extra credit as compensation. The subjects’ ages ranged from 18 to 30 years (mean 5 20.4 years, SD 5 2.7). None of the subjects reported having any allergies to certain tastes. The experimental procedures were approved by the Committee for the Protection of Human Subjects of Humboldt State University, and informed written consent was obtained from each subject. 2.1.2. Materials and procedures The five tastes used in the experiment were: apple juice (Tree Top Apple Juice, Tree Top Inc., Selah, WA), Gatorade (Lemon-Lime Gatorade, The Gatorade Company, Chicago, IL), water (Alhambra Mountain Spring Water, McKesson Water Products Co., Pasadena, CA), pickle juice (Del Monte Slow Cured Whole Pickle Juice, Del Monte Foods, San Francisco, CA), and soybean milk (Natural Unflavored Tofu Shop Soy Milk Specialty Foods Inc., Arcata, CA). Five plastic graduated syringes were used to measure the amount of liquid. The experiment was conducted in a private suite of a laboratory in the Department of Psychology. All 25 subjects had one laboratory session, in which they sat in a chair with a head rest. Two 6-mm silver–silver chloride cutaneous disk electrodes (Grass Instrument Division, Astro-Med, Inc., West Warwick, RI) were placed on the subjects’ left levator labii (superioris/alaeque nasi muscle) region to record facial EMG activity. Figure 1 illustrates the locations of two electrode placement. The levator labii superioris/alaeque nasi is
FIG. 1. Placement of electrodes on levator labii superioris alaeque nasi muscle region for surface EMG recordings.
a thin triangular muscle located by the side of the nose (Fig. 1). It begins by the inner margin of the orbit, and ends at upper lip and cartilage of the ala of the nose [10]. The electrodes were placed parallel to the muscle fibers with a distance of 6 mm between them. A third ground electrode was placed on the mastoid behind the left ear. Recording sites were prepared with rubbing alcohol and a mild skin abrasive. The electrodes were connected to an electromyogram amplifier model (EMG100B, Biopac Systems, Inc., Goleta, CA) with a frequency passband of 100–500 Hz. Then the EMG signals from the amplifier were digitized at a 2048 Hz sampling rate via a 16-bit analog/digital converter (UIM100, Biopac Systems, Inc., Goleta, CA) using a data acquisition and process software (MP100 Data Acquisition System, Biopac Systems, Inc., Goleta, CA) on an IBM-compatible Pentium computer. Each subject first sat still in the chair while recording 64 s of baseline EMG signals. Then, one of the experimenters infused 15 mL of the first liquid flavor into the subject’s mouth with a plastic syringe. The subject held the liquid in their mouth for 64 s while their facial EMG activities were recorded. Immediately after the liquid holding period, the subject read and made a single mark on a 100-mm horizontal visual-analog scale of palatability to the flavor. The scale ranged from 0 to 100 with “extremely like” as an anchor equal to 100 located at the right end of the scale, and “extremely dislike” as an anchor equal to 0 located at the left end of the scale. Then the subject spat out the liquid into a basin and rinsed their mouth thoroughly with water. Afterward, the subject rested for a period of 5 min. The subject repeated the same cycle of recording a baseline of EMG activity, recording a tasting period of EMG activity, marking the visual scale, and resting for 5 min after each of the remaining flavors. The five flavors tasted by each subject were: apple juice, Gatorade, water, soybean milk, and pickle juice. The order of the flavors was counterbalanced with a 5 3 5 Latin square design. 2.1.3. Data analysis The scores of palatability ratings on the visual scale were determined by measuring the length in millimeters from the left end of the scale to the mark. The mean scores for each flavor were then statistically compared. The EMGs were analyzed using the Fast Fourier Transform (FFT) spectral analysis technique (MP100 Data Acquisition system, Biopac Systems, Inc., Goleta, CA). Each 64 s of EMG data that was originally collected at the sampling rate of 2048 Hz during the baseline or flavor tasting periods delineated a time series epoch. After FFT analysis, the time series epoch was transformed into a frequency epoch with the range of 1 to 1024 Hz. Then, the numerical values were squared to obtain the spectral power, i.e., spectral density in mv2 for each frequency unit of the frequency epoch. The mean values of spectral power of the epoch were then obtained for the frequency band of 100 to 250 Hz. Because the numerical values were squared during the process
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
33
of obtaining spectral power, these values generated large variations between subjects. The ratio of EMG frequency band of 100 to 250 Hz between tasting and baseline periods was then calculated in the purpose of reducing variations in statistical analysis. 2.2. Results Table 1 presents the means and standard deviations for ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for the five different tastes (Table 1). As can be seen in Table 1, subjects reported the highest ratings of palatability to the apple juice, followed by Gatorade, water, soybean milk, and pickle juice. The subjects generated the lowest ratios of EMG spectral power between tasting and baseline periods of water, followed by apple juice, Gatorade, soybean milk, and pickle juice. The scatter plot of ratios of EMG spectral power against ratings of palatability when all data from five tasting conditions were pooled (Fig. 2) indicate a negative relationship (r 5 20.53, p , 0.0001). A univariate repeated measures analysis of variance on ratings of palatability indicated that there were significant differences among the tasting conditions, F(4, 96) 5 18.50, p , 0.001). Further univariate follow-up comparisons (paired t-tests) showed that the subjects during the apple juice tasting condition reported significantly higher ratings of palatability than during the soybean milk tasting condition, t(24) 5 4.65, p , 0.001, and than during the pickle juice tasting condition, t(24) 5 6.33, p , 0.001. The subjects during the Gatorade tasting condition reported significantly higher ratings of palatability than during the soybean milk tasting conditions, t(24) 5 2.72, p , 0.05, and than during the pickle juice tasting condition, t(24) 5 5.21, p , 0.001. The subjects during the water tasting condition reported significantly higher ratings of palatability than during the soybean milk tasting condition, t(24) 5 2.88, p , 0.01, and than during the pickle juice tasting condition, t(24) 5 5.63, p , 0.001. A univariate repeated-measures analysis of variance on ratios of facial EMG spectral power between tasting and baseline periods indicated that there were significant differences among the tasting conditions, F(4, 96) 5 8.79, p ,
Table 1 Means and standard deviations of ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for five different tastes Ratios of EMG spectral power
Ratings of palatability Flavors
n
Mean
SD
Mean
SD
Apple juice Gatorade Water Soybean milk Pickle juice
25 25 25 25 25
78.44 62.00 61.40 42.68 28.68
23.10 22.30 18.71 28.23 28.77
1.34 1.48 1.33 1.82 2.45
0.38 0.42 0.33 0.98 1.29
FIG. 2. Scatter plot of ratios of EMG spectral power versus ratings of palatability to the five flavors of 25 subjects.
0.001. Further univariate follow-up comparisons (paired ttests) showed that the subjects during the apple juice tasting condition generated significantly lower ratios of EMG spectral power between tasting and baseline periods than during the soybean milk tasting condition, t(24) 5 2.60, p , 0.05, and than during the pickle juice tasting condition, t(24) 5 4.14, p , 0.001. The subjects during the water tasting condition generated significantly lower ratios of facial EMG spectral power between tasting and baseline periods than during the soybean milk tasting condition, t(24) 5 2.30, p , 0.05, and than during the pickle juice tasting condition, t(24) 5 3.96, p , 0.001. 3. Experiment 2 3.1 Method 3.1.1 Subjects The same criteria and procedures in Experiment 1 were used in Experiment 2 to recruit 21 undergraduate students (6 men and 15 women). Their ages ranged from 18 to 30 years (mean 5 20.8 years, SD 5 1.9). 3.1.2. Materials and procedures The same equipment and procedures in Experiment 1 were used in Experiment 2. However, there were three tastes used in Experiment 2 that were: water (Alhambra Mountain Spring Water, McKesson Water Products Co., Pasadena, CA), a sugar solution: 20 g of sugar (Spreckels All Natural Sugar, Imperial Holly Corporation, Sugar Land, TX) dissolved in 150 mL water, and a salt solution: 5 g of salt (Morton Salt, Morton International Inc., Chicago, IL) dissolved in 150 mL water. The order of the tastes was counterbalanced with a 3 3 3 Latin square design.
34
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
3.2. Results Table 2 presents the means and standard deviations for ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for the three different tastes (Table 2). As can be seen in Table 2, subjects reported the highest ratings of palatability to water, followed by sugar and salt. However, the subjects generated the lowest ratios of EMG spectral power between tasting and baseline periods to water, followed by sugar and salt. The scatter plot of ratios of EMG spectral power against ratings of palatability when all data from three tasting conditions were pooled (Fig. 3) indicate a negative relationship (r 5 20.49, p , 0.001). A univariate repeated-measures analysis of variance on ratings of palatability indicated that there were significant differences among the tasting conditions, F(2, 40) 5 46.99, p , 0.001. Further univariate follow-up comparisons (paired t-tests) showed that subjects during the salt-tasting condition reported significantly lower ratings of palatability than during the water-tasting condition, t(20) 5 9.21, p , 0.001, and than during the sugar-tasting condition, t(20) 5 7.96, p , 0.001. No significant difference was found for ratings of palatability between the sugar and water tasting conditions. A univariate repeated-measures analysis of variance on ratios of facial EMG spectral power between tasting and baseline periods indicated that there were significant differences among the tasting conditions, F(2, 40) 5 10.70, p , 0.001. Further univariate follow-up comparisons (paired t-tests) showed that the subjects during the salt tasting condition generated significantly higher ratios of facial EMG spectral power between tasting and baseline periods than during the water tasting condition, t(20) 5 3.98, p , 0.001, and than during the sugar-tasting condition, t(20) 5 3.21, p , 0.01. No significant difference was found on ratios of facial EMG spectral power between tasting and baseline periods between water- and sugar-tasting conditions. 4. General discussion The present study investigated the sensory hedonic responses and associated facial EMG activities to different tastes. The results from the two experiments showed that
Table 2 Means and standard deviations of ratings of palatability and ratios of EMG spectral power between tasting and baseline periods for three different tastes Ratios of EMG spectral power
Ratings of palatability Flavors
n
Mean
SD
Mean
SD
Salt Sugar Water
21 21 21
22.57 67.00 72.76
21.81 18.56 24.17
1.87 1.31 1.25
0.79 0.58 0.41
FIG. 3. Scatter plot of ratios of EMG spectral power versus ratings of palatability to the water, sugar, and salt solutions of 21 subjects.
subjects’ ratings of palatability on the 0 to 100 hedonic scale ranged form 61.4 to 78.44 points while tasting apple juice, Gatorade, water, and sugar solution; and ranged from 22.57 to 42.68 points while tasting pickle juice, soybean milk, and salt solution. Because 50 points on the scale indicated a neutral hedonic level, these results suggest that apple juice, Gatorade, water, and sugar solution seem to generate positive hedonic responses, while pickle juice, soybean milk, and salt solution tend to generate negative hedonic responses. The experimental results also revealed that subjects generated higher ratios of facial EMG spectral power in the levator labii region between tasting and baseline periods to flavors with negative hedonic levels (such as pickle juice, soybean milk, and salt solution) than to flavors with positive hedonic levels (such as apple juice, Gatorade, water, and sugar solution). Thus, the negative hedonic sensations were associated with higher levator labii EMG activity, while the positive hedonic sensations were associated with lower levator labii EMG activity. These results suggest that EMG activity in facial levator labii muscles can be used as an indicator of palatability. The results of the present study are not only consistent with those in a recent study that showed disgust imagery is accompanied by increased EMG activities in the levator labii muscles [2], but may also support prior theoretical proposals that this particular muscle group has evolved as a functional necessity to protect the body from contamination, poisoning, or illness due to oral consumption [2,9,11,12]. It is known that the contraction of the levator labii muscle raises the upper lip and wrinkles the nose [2,8]. The contraction of the levator labii muscles occurs typically when it is necessary to avoid unpleasant odors by closing the nostrils and to expel noxious contents in the mouth by opening
S. Hu et al. / Physiology & Behavior 68 (1999) 31–35
the mouth. Previous research results from infant studies have also provided such theoretical support by noting that facial expressions of disgust existed in 2-day-old newborns when tasting solutions of citric acid, and that more positive facial expressions were observed when tasting solutions of sucrose [11]. The acts of using this muscle group surrounding the nose to detect or reject odoriferous foods and to open the mouth to dribble out unpalatable foods have been observed in both animal and human studies of taste reactivity and taste aversion [11,13]. However, it is important to point out that because present study recorded facial EMG activities from only one facial site, the levator labii muscle region, we are unable to say that the increase of EMG activity in the levator labii muscle region is the only indicator of palatability. To define the whole picture of facial EMG pattern associated with hedonic response, researchers need to record facial EMG signals from multiple sites in future sensory hedonic studies. Because previous studies have repeatedly demonstrated that laboratoryinduced unpleasant feelings are associated with high facial EMG activities in the corrugator regions [3,5,7,8], it is important to compare the differences between the EMG activities in the levator labii region and those in the corrugator region while tasting unpleasant flavors. Furthermore, the purpose of the present study was to investigate the relationship of palatability and facial EMG activities in the region of levator labii muscles. We were not attempting to define the facial EMG pattern of disgust. Because disgust is defined as a strong feeling of dislike or finding a thing very unpleasant [14], we consider that the negative hedonic sensation is related to the emotional state of disgust. However, it is too early to conclude that the increased EMG activity in the levator labii muscle region reflects only the emotional state of disgust. The eventual understanding of how disgust is manifested as an identified muscular pattern relies on the analysis of multiple facial EMG recordings from different facial regions. It is also impossible to use the data of the present study to verify other negative emotion states such as anger or sadness. We believe that negative hedonic experience is like an emotional experience that contains expressive, experiential, and physiological components [15–17]. These components are interrelated when emotional reactions are evoked [16– 19]. With the EMG recording, a physiological measure, researchers can obtain quantitative data, avoid human errors, and save time in coding facial motion patterns. Therefore, it
35
is recommended that EMG recording methods be used in future taste hedonic studies. References [1] Steiner JE, Lidar-Lifschitz D, Perl E. Taste and odor: reactivity in depressive disorders, a multidisciplinary approach. Percept Mot Skills 1993;77:331–1346. [2] Vrana SR. The psychophysiology of disgust; differentiating negative emotional context with facial EMG. Psychophysiology 1993;30:279– 86. [3] Cacioppo JT, Martzke JS, Petty RE, Tassinary LG. Specific forms of facial EMG response index emotions during an interview: from Darwin to the continuous flow hypothesis of affect-laden information processing. J Pers Soc Psychol 1988;54:592–604. [4] Dimberg U. Facial electromyography and emotional reactions. Psychophysiology 1990;27:481–94. [5] Schwartz GE, Ahern L, Brown SL. Lateralized facial muscle response to positive and negative emotional stimuli. Psychophysiology 1979;16:561–71. [6] Schwartz GE, Brown SL, Ahern L. Facial muscle patterning and subjective experience during affective imagery: Sex differences. Psychophysiology 1980;17:75–82. [7] Schwartz GE, Fair PL, Salt P, Mandel MR, Klerman GL. Facial muscle patterning to affective imagery in depressed and nondepressed subjects. Science 1976;192:489–91. [8] Cacioppo JT, Bush LK, Tassinary LG. Microexpressive facial actions as a function of affective stimuli: replication and extension. Pers Soc Psychol Bull 1990;18:515–26. [9] Vrana SR. Startle reflex response during sensory modality specific disgust, anger, and neutral imagery. J Psychophysiol 1994;8:211–18. [10] Gray H, Pick TP, Howden R. Gray’s Anatomy. Philadelphia, PA: Running Press, 1974. [11] Fox NA, Davidson RJ. Taste-elicited changes in facial signs of emotion and asymmetry of brain electrical activity in human newborns. Neuropsychology 1986;24:417–22. [12] Rozin P, Fallon AE. A perspective on disgust. Psychol Rev 1987;94: 23–41. [13] Parker LA. Rewarding drugs produce taste avoidance, but not taste aversion. Neurosci Biobehav Rev 1995;19:143–50. [14] Oxford American Dictionary. New York: Oxford University Press, 1980. [15] Izard CE, Kagan J, Zajonc RB. Emotions, Cognition, and Behavior. Cambridge, MA: Cambridge University Press, 1984. [16] Lang PJ. A bio-informational theory of emotional imagery. Psychophysiology 1979;16:495–512. [17] Zajonc RB. Feeling and thinking: preferences need no inferences. Am Psychol 1980;35:151–75. [18] Ekman P. Cross-cultural studies of facial expressions. In: Ekam, P, editor. Darwin and Facial Expression. New York: Academic Press, 1973, pp. 169–222. [19] Lazarus RS. Thoughts on the relations between emotion and cognition. In: Scherer KL, Ekman P., editors. Approaches to Emotion. Hillsdale, NJ: Erlbaum, 1984, pp. 247–57.