- Email: [email protected]
Autonomic, neuroendocrine, and subjective responses to emotion-inducing film stimuli
International Journal of Psychophysiology, 11 (1991) 131-140 ?I 1991 Elsevier Science Publishers B.V. 0167-8760/91/$03.50
PSYCHO
131
00342
Autonomic,
neuroendocrine, and subjective to emotion-inducing film stimuli Walter Hubert
responses
and Renate de Jong-Meyer
Institute of Psychology I, Unwersity of Miinster, D-4400 Miinster (17 R.G.) (Accepted
Key words:
Emotion;
Film; Mood;
Bodily Sensation;
1 October
Interoception;
1990)
Heart rate; Skin conductance;
Salivary
cortisol
The aim of the present study was to investigate differential emotional response patterns to film stimuli by multimodal assessment of subjective, autonomic and endocrine variables. Scenes of two films, chosen to induce different affective states, were presented to 20 male subjects. Changes in mood and perceived bodily sensations, autonomic activity (heart rate and skin conductance) and one endocrine parameter (salivary cortisol) were measured. Analyses revealed that the films elicited differential mood patterns. The cartoon film, which induced a pleasant, amused state, was accompanied by very few changes in bodily sensations, a temporary decrease in heart rate, and a rapid decrease in electrodermal activity. The suspense film, which elicited a marked reduction in joyfulness and relaxation together with an enhancement in irritation, differed from that pattern primarily by marked changes in several bodily sensations, an increase in electrodermal activity, and a temporary decrease in heart rate. There were no film-dependent cortisol changes. The results together with evidence from other emotion induction experiments support the usefulness of film segments in eliciting mild to moderate affective states and their concomitances. They particularly point to the relevance of bodily sensations as an emotional response component.
INTRODUCTION The use of psychological indices as a measure of emotion is based on the assumption that different emotions are associated with distinctive patterns of physiological response. Some theorists have argued that all emotions are based on a state of non-specific arousal (e.g., Cannon, 1927; Schachter and Singer, 1962), and that an emotional state is the result of the interaction between physiological arousal and a cognition about the arousing situation. Others, however, have stated that emotions have specific physiological response patterns (e.g., James, 1890; Ax, 1953; Ekman et al., 1983; Schwartz et al., 1981; Mason, 1975;
Correspondence: sity of Miinster,
W. Hubert, Institute Rosenstr. 9, D-4400
of Psychology I, UniverMiinster, F.R.G.
Lang, 1979; Lang et al., 1980). There is considerable evidence for the latter view indicating that autonomic activity is differentiated along multiple evaluative dimensions of emotion (e.g., Ax, 1953; Ekman et al., 1983; Schwartz et al., 1981; Winton et al., 1984; Lang et al., 1980) although the patterns of autonomic activity associated with different emotions have not yet been clearly identified. Changes in heart rate (HR) are innervated by both the sympathetic and parasympathetic systems, and it is assumed that HR is physiologically and psychologically determined (Lacey et al., 1963; Obrist, 1976, Fowles, 1980). Electrodermal activity (EDA), on the other hand, is innervated entirely by the sympathetic system and may unambiguously reflect primary arousal. Future research is required to investigate changes in HR and EDA and their potential interaction to distinct emotional stimuli.
132
Since it had been difficult to demonstrate a clear relationship between specific physiological processes and specific emotions, the question arised; how bodily sensations relate to physiological changes and both to subjective emotional experiences. Some studies demonstrated a differential perception of bodily sensations in association with different recalled emotional experiences (e.g., Fahrenberg, 1965; Landy and Stern, 1971; Shields, 1984; Scherer and Tannenbaum, 1986). But acute changes in perceived bodily sensations were rarely assessed (e.g., Tyrer et al., 1978; de Jong-Meyer et al., 1990). All three relevant response components (physiological recordings, self-report of mood and perceived bodily sensations) were investigated in a few studies only (e.g., Pennebaker, 1982; de JongMeyer et al., 1990). From these studies it can be concluded, that relationships between objectively recorded physiological changes and perceived changes show a very high interindividual variability and are generally low. Relationships between bodily sensations and affective states, on the other hand, are more consistent and show some specificity, particularly regarding unpleasant valenced states like anxiety, sadness or anger. Several findings support the conclusion that emotional experiences are accompanied by changes in the endocrine system (for review see Brady, 1975; Mason, 1975; Henry, 1986). Bermond and Frijda (1987) have argued that endocrine feedback mechanisms may regulate the intensity of emotional experiences. Emotion-induced changes in the level of hormones, often a relatively slow process, may not have direct effects on the emotional experience, but may predispose the individual to react in a certain way during future emotion eliciting situations. Whereas important in endocrine stress research (see review by Rose, 1984) the pituitary-adrenalcortical system has rarely been investigated in the area of differential emotion research. More generally, anxiety (e.g., Bloch and Brackenridge, 1972; Brady, 1975; Brown and Heninger, 1975; Francis, 1979; Hubert and de Jong-Meyer, 1989; Hubert et al., 1989) depression (e.g., Sachar, 1981) as well as bereavement and sadness (e.g., Hofer et al., 1972a, b; Jacobs et al., 1985; Irwin et al., 1988) have been associated with an elevation of cortisol, while re-
laxation and meditation on the other hand were found to correlate with declines in cortisol levels compared to control conditions (e.g., Jevning et al., 1978, DeGood and Redgate, 1982). Systematic research on positive emotions, for example joy or happiness, are lacking. Research on endocrine response patterns during and following psychological stimuli has often been restricted by the collection of blood samples. The possibility of assessing hormones in saliva is of great advantage, since saliva samples are easily and repeatedly obtainable and do not demand stressful venipuncture (see review by Riad-Fahmy et al., 1982). Fortunately, saliva flow-rate does not influence salivary cortisol levels. It could be demonstrated that there is an excellent correlation (r > 0.90) between serum free and saliva cortisol levels. Therefore, salivary cortisol measurements would appear to be the method of choice in human psychoendocrinological research (Bassett et al., 1987). Film stimuli are often used as elicitors of mood changes. McHugo et al. (1982) for example, demonstrated that film segments can reliably elicit both pleasant and unpleasant valenced affective changes. Across films a wide range of affective responses can be induced, thus film segments represent one rich source of stimuli for emotional research. The present film study was designed to induce changes in affective states, which were assumed to be accompanied by differential response patterns in perceived bodily sensations, autonomic activity and salivary cortisol. To get differential changes in affect, film content was chosen to evoke a pleasant, amusing versus an unpleasant, arousing state of suspense. Suspense, not unpleasantness per se, was shown in prior film studies to elicit cortisol changes (Hellhammer et al., 1986). The other assumptions, based on previous film and other mood induction results were, that perceived bodily changes would be enhanced after the suspense scenes, that electrodermal activity would be increased during the suspense and decreased during the pleasant, amusing film because of the different arousal impact, and that the HR deceleration (because of the shared information intake type of situation, Lacey et al., 1963) would be reduced
133
during the suspense film by stimuli high in preparedness (Cook et al., 1986) potential for avoidance tendencies (snakes, spiders).
MATERIALS
AND
METHODS
Subjects The subjects were 20 healthy male students (excluding psychology students) aged 21-24 years (M = 22.7 years). Written informed consent was obtained from all. They were paid DM 50 for participation in the experiment. Subjective rating scales Mood ratings were assessed using eight visual analogue scales (VAS; Aitken, 1969; see Fig. 1 for content). A list of bodily sensations consisted of 14 VAS items (see Fig. 2 for content). For each item the subject had to state on a lOO-mm line (0: not at all, 100: very much) to what extent he experienced the corresponding mood/bodily sensation at the moment. These items were shown in previous studies to discriminate between different affective states (Hubert and de Jong-Meyer, 1989; de Jong-Meyer et al., 1990). Apparatus HR (bpm) was recorded by conventional ECG (lead II) using Ag/AgCl electrodes filled with ECG paste. ECG signal was amplified by Natic UBV300 and the R-R intervals were scored and converted to HR using a Kontron PSI980 computer. EDA was recorded by a constant voltage of 0.5 V with Beckman Ag/AgCl electrodes (8 mm diameter) filled with 0.05 M NaCl Unibase electrode paste. The recording was taken from thenar-hypothenar sites of the left hand using a Natic EDA coupler. The original DC signal (skin conductance level, SCL, pS) and the AC signal with a time constant of 10 s were sampled at 10 Hz and recorded on disk. Criterion for spontaneous fluctuations (SFs, rate/mm) in skin conductance was 2 0.1 pS. Salivary cortisol Subjects collected l-2 ml saliva at -15, 0, 15, 30 and 45 min referring to the onset of the film.
Saliva samples were stored at - 20” C until assayed. Salivary cortisol was determined in duplicate by using a commercial radioimmunoassay kit from Mallinckrodt Diagnostica (SPAC Cortisol, Dietzenbach, F.R.G.). The intra- and interassay coefficients of variation were 1.9-2.3% and 8.4%, respectively. Procedure Two lo-min film segments were used as emotional stimuli. They were presented in counterbalanced order on consecutive days (one film per day). The suspense segment from ‘Indiana Jones Raiders of the lost Ark’ (U.S.A. 1980, directed by S. Spielberg) showing also situations with snakes and spiders were chosen as the unpleasant, arousing stimulus. Scenes from ‘Peanuts’ (TV program) were chosen as the pleasant and amusing stimulus. The film segments had been rated by a different sample of subjects as eliciting a calm, joyful or an unpleasant, arousing mood, respectively. In an introductory session (1 h) subjects got familiarized with the procedure and equipment. The two experimental sessions (1.5 h) were conducted in a light-controlled and soundproof laboratory, with the subject seated in a comfortable semi-reclining chair in front of a TV monitor. Subjects were instructed to refrain from alcohol or other drugs and exercise prior to the sessions. To control for circadian variations of salivary cortisol all sessions were conducted between 13.00 and 14.30 or between 15.00 and 16.30 h (randomly assigned). Sessions followed the same schedule: (a) attachment of electrodes, (b) 3-min baseline recording, (c) lo-min film presentation, (d) 35-min post film period. Onset of the film was 30 min after arrival. Physiological recordings were done continuously during baseline and film presentation. Mood and bodily sensations were recorded before and following film presentation. Data scoring and analysis The values of HR and SCL were reduced to one average for 60-s periods. For SFs time period was also 60-s. For the physiological recordings, mean values of the last 2 min of the baseline period were used as baseline scores. For cortisol the baseline levels were at time 0-min.
134
For analyses of variance (ANOVA) and multivariate analyses of variance (MANOVA), Greenhouse-Geisser corrections on the degrees of freedom were applied where appropriate.
Mood Item
Tired
RESULTS
Relaxed
I
_I
-
1
i-r--
-
~-~
~ Mood Differences among baselines of all mood items for the two films were not significant. A 2 (Film) x 2 (Time) repeated-measures MANOVA was performed for all mood items as dependent variables. It revealed a significant Time main effect (F&11) = 4.17, P = 0.016) and a significant Film x Time interaction (F&11) = 8.40. P < 0.001). The main effect for Film was not significant. Follow-up univariate ANOVAs of the significant multivariate Time main effect and multivariate Film x Time interaction produced significant Time main effects for ‘relaxed’ (F(1,9) = 22.76, P = 0.001) and ‘curious’ (F(1,9) = 19.22, P = 0.002), while the Film x Time interaction was significant for ‘relaxed’ (F(l,S) = 8.02, P = 0.022), ‘joyful’ (F(1,8) = 22.09, P = 0.002) and ‘irritated’ (F(1,8) = 16.00, P = 0.004). As illustrated in Fig. 1, there were increases for ‘irritated’ (F(1,6) = 9.66, P = 0.021), and decreases for ‘relaxed’ (F(1,6) = 24.44, P = 0.003) and ‘joyful’ (F(1,6) = 11.14, P = 0.015) during the suspense scenes. On the other hand, the pleasant, amusing film segment induced an increase for ‘joyful’ (F(1,9) = 10.94, P = 0.009). There was a decline for ‘curious’ after both films (negative: F(1,9) = 21.80, P = 0.001; positive: F(1,9) = 7.93, P = 0.020). Bodily sensations Differences among baselines of all bodily sensation items for the two films were not significant. A 2 (Film) x 2 (Time) repeated-measures MANOVA was performed for all bodily sensation items as dependent variables. It revealed a significant Film X Time interaction (F(I4,5) = 5.51, P = 0.035). The main effects for Film or Time were not significant.
Joyful
Sad
Irritated
Curious
Vigorous
Anxious
i
i_ ~.
-1
-30
-20
-A----10
0
~~_~ 10
20
30
VAS-Diff. m Fig. 1. Mean
pleasant/amuwng
r=-i unpleasant/arousing
changes in mood during pleasant, unpleasant, arousing film stimulation.
amusing
and
Follow-up univariate ANOVAs of the significant multivariate Film X Time interaction produced significant Film X Time interactions for ‘hot face’ (F(1,4) = 8.56, P = 0.043), ‘sweating’ (F(1,4) = 9.26, P = 0.038), ‘heart rate increasing’ (F(1,4) = 17.94, P = 0.013), ‘sweaty palms’ (F (1,4) = 28.39, P = 0.006), ‘restless’ (F(1,4) = 10.87, P = 0.030) and ‘difficulties in breathing’ (F(1,4) = 12.58, P = 0.024). As illustrated in Fig. 2, there were significant increases for ‘hot face’ (F(1,4) = 10.25, P = 0.033), ‘sweating’ (F(1,4) = 15.18, P = 0.017), ‘heart rate increasing’ (F(1,4) = 23.37, P = 0.008), ‘sweaty palms’ (F(1,4) = 32.68, P = O.OOS), ‘restless’ (F(1,4) = 20.26, P = 0.010) and ‘difficulties in breathing’ (F(1,4) = 11.46, P = 0.027) during the suspense scenes, while during the pleasant, amusing film ‘restless’ (F(1,4) = 7.76, P = 0.049) increased significantly.
135
11.70, P < 0.001) and SFs (F(6,116) = 6.69, P < 0.001). Multiple t-tests were performed to compare mean changes during film stimulation to baseline levels. HR decreased from minutes 2 to 4 (for all P -Z 0.05) during the suspense scenes and from minutes 2 to 5 (for all P < 0.05) during the pleasant, amusing film segment (Fig. 3). For SCL there were significant increases for minutes 3 and 6 (both P < 0.01) during the suspense scenes, and after an initial increase for minute 1 (P < 0.01) there were significant decreases from minutes 5 to 10 (for all P < 0.05) during the amusing film (Fig. 4). SFs increased significantly for minute 1 and for minutes 3 to 10 (for all P < 0.05) during the suspense film and decreased insignificantly after an initial increase for minute 1 (P < 0.05) for minutes 3 to 8 during the amusing film (Fig. 5).
Bodily sensations item Hot face Feeling
cold
Sweating Shallow
breathing
HR increasing Breathing
faster
Blood rush. to head Tense
atomaoh Trembling
Close Paralyaed
to tears with fear
Sweaty
palm8
Ffeatleasnesa Diff.
in breathing
HR
VAS-Diff. m
positive
Fig. 2. Mean changes in the perception during pleasant, amusing and unpleasant, lation.
Autonomic
negative of bodily sensations arousing film stimu-
activity
Differences among baselines for the two films were not significant. A 2 (Film) x 11 (Time) repeated-measures multivariate analysis of variance (MANOVA) was performed for HR, SCL, and SFs as dependent variables. It revealed significant main effects for P = 0.016) and Time Film (1;(1,6) = 10.90, (F(6,108) = 4.94, P < 0.001). The Film X Time interaction was also significant (F(5,85) = 5.25, P < 0.001). Follow-up univariate ANOVAs produced significant main effects for Film on the measures of SCL (F(1,19) = 9.58, P = 0.006) and SFs (F(1,19) = 34.14, P < O.OOl), and for Time on the measures of HR (F(4,74) = 7.07, P = 0.002), SCL (F(2,46) = 4.04, P = 0.024) and SFs (F(6,116) = 3.79, P < 0.001). Furthermore, the Film X Time interaction was significant for SCL (P(2,46) =
73
71 c
69t
67
1 7%
66’
-’
0
+
pleasant/amusing
’ 1
’ 2
3 Time
Fig. 3. Mean changes amusing and unpleasant,
4
unpleasant/arousing
,
1
5
6
in 60-s
1-d 7
6
9
IO
intervalls
in heart rate (HR) during pleasant, arousing film stimulation (0 indicates baseline).
136
Salivury cortisol Differences among baselines for the two films were not significant. A 2 (Film) X 5 (Time) repeated-measures ANOVA for salivary cortisol levels revealed a significant main effect for Time (F(2,30) = 56.96, P < 0.001). The main effect for Film and the Film X Time interaction were not significant. As illustrated in Fig. 6 salivary cortisol levels declined during both conditions.
l/min 61
DISCUSSION The aim of this study was to use multimodal assessment of subjective, autonomic and endocrine variables to investigate differential responding to emotion-inducing film segments. 1
’ +
pleasantlamuslng
0
SCL
0
1
2
3
++
1
1
4
5
unpleasant/arousing
I 6
7
8
9
IO
Time in 60-s intervalls Fig. 5. Mean changes in spontaneous fluctuations of skin conductance (SFs) during pleasant. amusing and unpleasant, arousing film stimulation (0 indicates baseline).
te-
+ 0
if
pleasant/amusing
1
2
3
4
unpleasant/arousing
1
1
1
1
5
6
7
8
‘~.
9
10
Time in 60-s intervalls Fig. 4. Mean responses in skin conductance level (SCL) during pleasant, amusing and unpleasant, arousing film stimulation (0 indicates baseline).
The results showed that film segments induced an unpleasant, irritated and arousing versus joyful state. Mood profiles differed clearly and thus showed, that film stimuli can be used for differentiated mood induction purposes, and are in line with similar findings (e.g., Lazarus et al. 1962; Tyrer et al., 1978; McHugo et al., 1982; Hubert and de Jong-Meyer, 1989). The evoked changes in perceived bodily sensations supported our assumptions. After the suspense scenes subjects reported changes in temperature related sensations (hot face, sweating, sweaty palms), difficulties in breathing, an increased HR and more restlessness, while the amusing film elicited a small increase in restlessness and no changes in other bodily sensations. The results are in line with Tyrer et al. (1978), Pennebaker (1982) and de Jong-Meyer et al. (1990). Furthermore, these findings expand results of Fahrenberg (1965).
137
Salivary
cortisol
18
16, ‘9
14.
12
87 + 6 I I___~~ -15
pleasant/amusing
+
unpleasantlarousfng
1
I
0
15
Time
30
45
[mid
Fig. 6. Mean changes in salivary cortisol before and following pleasant, amusing and unpleasant, arousing film stimulation.
Landy and Stern (1971), Shields (1984) and Scherer and Tannenbaum (1986), who demonstrated differential perceptions of bodily sensations in association with different recalled emotional experiences. The results point to the importance of assessing perceived bodily sensations as a component of emotional responding, particularly, if one considers the only mild to moderate range of reported mood changes. Future studies should include the assessment of bodily sensations to decide, whether other affective states are accompanied by distinct profiles of these sensations, and whether they are indicative of specific unpleasant emotional states or of different bodily arousal patterns. The finding, that pleasant and/or relaxed states are characterized by a lack of changes in bodily sensations, if replicated, has important implications for the elaboration of treatment strategies for psychosomatic or affective disorders.
On the autonomic level there were differential response patterns for SCL and SFs. The significant increase during the first minute of both films may be interpreted as an orienting response. The increases in SCL and SFs during the suspense scenes are similar to results of Lazarus et al. (1962) Mordkoff (1964) Kaiser and Roessler (1970) and Steptoe and Vbgele (1986) using also unpleasant and arousing film segments. The decrease in SCL and SFs during the amusing scenes were found in a study of Kaiser and Roessler (1970) too. Regarding HR, we found decelerations for both film segments and no differences between them. This is in line with the suggestion of Lacey et al. (1963) that situations with the intake of informations are characterized by HR decelerations. We had assumed, however, that because of the high potential of our suspense film to provoke avoidance tendencies, HR deceleration would be reduced during this film, leading to differences compared to the amusing one. Although the relevant Film X Time interaction failed to become significant, the results do not argue strongly against this assumption. Subjectively, the participants got more irritated, and there were stronger HR accelerations during the second part of the suspense segment, while during the amusing film this trend was weaker towards the end. The literature regarding the effects of film stimuli on HR is mixed. Whereas Lazarus et al. (1962) Mordkoff (1964) and Steptoe and Vbgele (1986) reported significant increases of HR during the presentation of unpleasant, arousing film stimuli, Carruthers and Taggart (1973) demonstrated a HR deceleration as a response to a violent movie, even in the presence of increased sympathetic activity. Van den Akker and Steptoe (1987) could not demonstrate an acceleration in HR during an industrial accident film. Further studies should test the functional significance of HR changes in relation to the information intake versus rejection assumption. They should allow for more objective control of this information processing dimension and then vary the valence of the elicitor. The expected increase in salivary cortisol levels in response to suspense could not be observed.
13X
During both conditions there were decreases in salivary cortisol levels. Cortisol in serum and saliva has a marked circadian rhythm (Riad-Fahmy et al., 1982). During morning and afternoon hours cortisol levels decrease continuously. This circadian rhythm seems to be the reason for the decrease in salivary cortisol levels in this study. One may argue that the time course of saliva sampling is not sufficient to detect increases in salivary cortisol levels. But Riad-Fahmy et al. (1983) demonstrated that time intervals of 15 min provide a convenient way to estimate salivary cortisol variations. Another reason for the lacking activation of the adrenal system during suspense might be the short duration of the film and/or its mild to moderate affective changes. Other studies demonstrating a significant cortisol elevation used longer films of 30 to 90 min duration (e.g., Bernick et al., 1971; Brown and Heninger, 1975; Hellhammer et al., 1986; Bossert et al., 1988; Hubert and de Jong-Meyer, 1989) and stress of venipuncture, a was shown to be effective in strong stimulus, inducing cortisol elevations (Mason et al.. 1973; Rose and Hurst, 1975; Hubert et al., 1989). Salivary cortisol declined more rapidly prior film viewing compared to the period during and following film presentation. One may argue that the setting up procedures for the experiment were stressful and induced a cortisol release with a subsequent rapid decrease. But inspection of the mood and bodily sensation ratings at baseline do not support this argumentation. One explanation for this initial rapid decrease might be that physical exercise (e.g., Cook et al.. 1987; O’Connor and Corrigan, 1987) produced an elevation in cortisol level while subjects came to the laboratory. The first saliva sample was assessed already 15 min after arrival. Future studies should extent the time of adaptation in order to minimise such effects. Taken together, our results as well as those from other film studies prove this paradigm to be useful in eliciting mild to moderate affective states and their concomitants. Further investigations should include more different emotions, particularly distinct pleasant and unpleasant ones, varying in degree of arousal, and should control attentional set and direction of information processing. Differential perceived bodily sensations as well as
the contribution of HR changes to differential emotion profiles seem to be promising. Furthermore, psychoendocrinological investigations should use longer film segments to study cortisol response patterns during different emotions. Given that subsequent studies with a larger sample of males and females replicate and extend these findings we may learn more on pattern classification of different emotions.
ACKNOWLEDGMENTS This study was supported partly by a grant Forschungsgemeinschaft (Jo from the Deutsche 140/l -2). The authors thank Prof. E. Nieschlag of Reproductive and Mrs. M. Miiller (Institute University of Miinster) for cortisol Medicine, analyses.
REFERENCES Altken. R.C. (1969) A growing edge of measurement of feelings. Proc. Roy. SW. Med.. 62: 989-993. Ax. A.F. (1953) The physiological differentiation between fear and anger in humans. Pq&xon~. Med., 15: 433-442. Bassett, J.R., Marshall, P.M. and Spillane, R. (1987) The physiological measurement of acute stress (public speaking) in bank employees. Inr. J. Psychoph~vsiol., 5: 265-213. Bermond, B. and Frijda, N.H. (1987) The intensity of emotional feelings: Product of peripheral emotional responses or cognitions? Commun. Cogn., 20: 191-205. Bernick. N., Kling, A. and Borowitz, G. (1971) Physiologic differentiation of sexual arousal and anxiety. P.~ychmom. Med.,
33: 341-352.
Bossert, S., Berger, M.. Krieg, J, C., Schreiber, W.. Junker, M. and von Zerssen, D. (1988) Cortisol response to various stressful situations: Relationship to personality variables and coping styles. Neuropsychohiology. 20: 36-42. Brady, J.V. (1975) Toward a behavioral biology of emotion. In L. Levi, (Ed.), Emorron.s Their prrramerer.~ and memmment. Raven Press. New York, pp. 17-45. Brown, W.A. and Heninger, G. (1975) Cortisol. growth hormone. free fatty acids. and experimentally evoked affective arousal. Am. J. P.ychmr.. 132: 1172-1176. Cannon, W.B. (1927) The James-Lange theory of emotion: A critical examination and an alternative theory. Ant. J. Psychol., 39, 106-124. Carruthers, M. and Taggart, P. (1973) Vagotonicity of violence: Biochemical and cardiac responses to violent film5 and television programmea. Bnr. Med J., 18: 384-389.
139 Cook III, E.W., Hodes, R.L. and Lang, P.J. (1986) Preparedness and phobia: Effects of stimulus content on human visceral conditioning. J. Ahn. Psychoi., 95: 195-207. Cook. N.J., Ng, A., Read, G.F., Harris, B. and Riad-Fahmy, D. (1987). Salivary cortisol for monitoring adrenal activity during marathon runs. Harm. Res., 25: 18-23. DeGood, D.E. and Redgate, E.S. (1982) Interrelationship of plasma cortisol and other activation indices during EMG biofeedback training. .I. Behau. Med., 5: 213-223. De Jong-Meyer. R., Hubert, W., KGrner, U. and Geyer. E. (1990) Induktion von Angst and Freude fiber die Imagination autobiographischer Erinnerungen [The induction of anxiety and happiness by the imagery of autobiographical recollection]. 2. Klin. Psychol., 19: 83-95. Ekman, P.. Levenson, R.W. and Friesen, W.V. (1983) Autonomic nervous system activity distinguishes among emotions. Science, 221: 1208-1210. Fahrenberg, J. (1965) Zur Frage einer differentiellen Physiologie der Affekte [Questions about a differential physiology of affects]. Psycho. Forschung, 28: 422-438. Fowles. D.C. (1980) The three arousal model: Implications of Gray’s two-factor learning theory for heart rate, electrodermal activity, and psychopathy. Prychophysmlog);. 17: X7-104. Francis. K.T. (1979) Psychologic correlates of serum indicators of stress in man: A longitudinal study. Ps)?chosom. Med., 41: 617-628. Hellhammer, D.H., Rottger, K.. Lorenzen, J. and Hubert, W. (1986) Suspense increases salivary cortisol. In R.C. Fredrickson, H.C. Hendrie. J.N. Hingtgen and M.H. Aprison (Eds.), Neuroregulation of au~onomrc, endocrine and immune systems, Martinus Nijhoff, Boston, pp. 521-524. Henry, J.P. (1986) Neuroendocrine patterns of emotional response. In R. Plutchik and K. Kellerman (Eds.), Emorion Theory, research, and experience. Vol. 3, Biological foundations of emotion, Academic Press, Orlando, pp. 37-60. Hofer. M.A., Wolff, C.T., Friedman, S.B. and Mason, J.W. (1972a) A psychoendocrine study of bereavement. Part I. 17-hydroxycorticosteroid excretion rates of parents following death of their children from leukemia. Psychosom. Med.. 34: 481-491. Hofer. M.A., Wolff, C.T., Friedman, S.B. and Mason, J.W. (1972b) A psychoendocrine study of bereavement, Part Il. Observations on the process of mourning in relation to adrenocortical function. Psychosom. Med., 34: 492-504. Hubert, W. and de Jong-Meyer, R. (1989) Emotional stress and saliva cortisol response. J. Clin. Chem. C/in. Biochem., 27: 235-237. Hubert, W., Miiller, M. and Nieschlag, E. (1989) Stress reactions in response to the procedure of LHRH-tests as measured by salivary and serum cortisol and psychological variables. Harm. Res., 32: 198-202. Irwin. M., Daniels, M., Risch, SC., Bloom, E. and Weiner, H. (1988) Plasma cortisol and natural killer cell activity during bereavement. Biol. Psychiat.. 24: 173-178. Jacobs. S.. Mason, J., Kosten, T., Kasl, S., Ostfeld, A.. Atkins,
S., Gardner, C. and Schreiber, S. (1985) Acute bereavement, threatened loss, ego defenses and adrenocortical function. Psychother. Psychosom.. 44: 151-159. James, W. (1890) The principles OJ psychology, Henry Holt, New York. Jevning, R., Wilson, A.F. and Davidson, J.M. (1978) Adrenocortical activity during meditation. Harm. Behau., 10; 54-60. Kaiser, C. and Roessler, R. (1970) Galvanic skin responses to motion pictures. Pert. Mot Sk.. 30: 371-374. Lacey, J.I., Kagan, J., Lacey, B.C. and Moss, H.A. (1963) The visceral level: Situational determinants and behavioral correlates of autonomic response patterns. In P. Knapp (Ed.), Expressions of the emotions in man. International University Press, New York. pp. 161-196. Landy. F.J. and Stern, R.M. (1971) Factor analysis of a somatic perception questionnaire. J. Psychosom. Res.. 15: 179-181. Lang, P.J. (1979) A bio-information theory of emotional imagery. Psychoph_v.volo~. 16: 495-512. Lang, P.J., Kozak, M.J., Miller, G.A.. Levin, D.N. and McLean, A. (1980) Emotional imagery: Conceptual structure and pattern of somato-visceral response. Psychophysrology, 17: 179-192. Lazarus. R.S., Speisman, J.C., Mordkoff, A.M. and Davison, L.A. (1962) A laboratory study of psychological stress produced by a motion picture film. Psychol. Monogr., 76: Whole No. 553. Mason, J.W. (1975) Emotion as reflected in patterns of endocrine integration. In L. Levi (Ed.), Emotions - Their parameters and measurement, Raven Press. New York, pp. 1433 181. Mason. J.W., Hartley, L.H.. Kotchen, T.A., Mougey, E.H., Ricketts, P.T. and Jones, L.G. (1973) Plasma cortisol and norepinephrine responses m anticipation of muscular exercise. Psychosom. Med.. 35: 406-414. McHugo, G.J.. Smith, C.A. and Lanzetta. J.T. (1982) The structure of self-reports of emotional responses to film segments. Motiu. Emot.. 6: 365-385. Mordkoff. A.M. (1964) The relationship between psychological and physiological response to stress. Psychosom. Med.. 26: 135-150. Obrist, P.A. (1976) The cardiovascular-behavioral interaction As it appears today. Psychophysiology, 13: 95-107. O’Connor, P.J. and Corrigan. D.L. (1987) Influence of shortterm cycling on salivary cortisol levels. Med. SCL Sports Exert., 19: 224-228. Pennebaker, J.W. (1982) The psychology of physical symptoms. Springer-Verlag, New York. Riad-Fahmy, D., Read. G.F. and Walker, R.F. (1982) Steroids in saliva for assessing endocrine function. Endocr. Rev.. 3: 367-395. Riad-Fahmy, D., Read, G.F. and Walker. R.F. (1983) Salivary steroid assays for assessing variation in endocrine activity. J. Steroid Biochem., 19: 265-272. Rose, R.M. (1984) Overview of endocrinology of stress. In G.M. Brown, (Ed.). Neuroendocrmology cmdpsychmtric disorder, Raven Press. New York. pp. 95-122.
140 Rose. R.M. and Hurst. M.W. (1975) Plasma cortisol and growth hormone responses to intravenous catheterization. J. Human Stress, 1: 22-36. Sachar, E.J. (1981) Psychobiology of affective disorders. In E.R. Kandel and J.H. Schwartz (Eds.), Principles of neural .scrence, Elsevier. New York, pp. 61 I-61 9. Schachter, S. and Singer, J. (1962) Cognitive, social, and physiological determinants of emotional state. Psycho/. Ret:.. 69, 379-399. Scherer, K.R. and Tannenbaum. P.H. (1986) Emotional experiences in everyday life: A survey approach. Mob. Emot., 10: 295-314. Schwartz. G.E.. Weinberger. D.A. and Singer, J.A. (1981) Cardiovascular differentiation of happiness. sadness, anger. and fear following imagery and exercise. P.rycho.wm. Med.. 43: 343-364.
Shields, S.A. (1984) Reports of bodily change in anxiety, sadness, and anger. Motic,. Emot., 8: l-20. Steptoe, A. and VGgele, C. (1986) Are stress responses influenced by cognitive appraisal? An experimental comparison of coping strategies. Brrt. J. PsychoI.. 71: 243-255. Tyrer. P.. Lewis. P. and Lee. I. (1978) Effects of subliminal and supraliminal stress on symptoms of anxiety. J. Nem. Ment. DlS., 166: 88-95. Van den Akker, 0. and Steptoe, A. (1987) Psychophysiological responses in women with premenstrual and menstrual symptoms. J. Psychoph_ysiol., 1: 149-158. Winton. W.M., Putnam. L.E. and Krauss, R.M. (1984) Facial and autonomic manifestations of the dimensional structure of emotion. J. Exp. Sot. Ps_vchol.. 20: 195-216.
PSYCHO
131
00342
Autonomic,
neuroendocrine, and subjective to emotion-inducing film stimuli Walter Hubert
responses
and Renate de Jong-Meyer
Institute of Psychology I, Unwersity of Miinster, D-4400 Miinster (17 R.G.) (Accepted
Key words:
Emotion;
Film; Mood;
Bodily Sensation;
1 October
Interoception;
1990)
Heart rate; Skin conductance;
Salivary
cortisol
The aim of the present study was to investigate differential emotional response patterns to film stimuli by multimodal assessment of subjective, autonomic and endocrine variables. Scenes of two films, chosen to induce different affective states, were presented to 20 male subjects. Changes in mood and perceived bodily sensations, autonomic activity (heart rate and skin conductance) and one endocrine parameter (salivary cortisol) were measured. Analyses revealed that the films elicited differential mood patterns. The cartoon film, which induced a pleasant, amused state, was accompanied by very few changes in bodily sensations, a temporary decrease in heart rate, and a rapid decrease in electrodermal activity. The suspense film, which elicited a marked reduction in joyfulness and relaxation together with an enhancement in irritation, differed from that pattern primarily by marked changes in several bodily sensations, an increase in electrodermal activity, and a temporary decrease in heart rate. There were no film-dependent cortisol changes. The results together with evidence from other emotion induction experiments support the usefulness of film segments in eliciting mild to moderate affective states and their concomitances. They particularly point to the relevance of bodily sensations as an emotional response component.
INTRODUCTION The use of psychological indices as a measure of emotion is based on the assumption that different emotions are associated with distinctive patterns of physiological response. Some theorists have argued that all emotions are based on a state of non-specific arousal (e.g., Cannon, 1927; Schachter and Singer, 1962), and that an emotional state is the result of the interaction between physiological arousal and a cognition about the arousing situation. Others, however, have stated that emotions have specific physiological response patterns (e.g., James, 1890; Ax, 1953; Ekman et al., 1983; Schwartz et al., 1981; Mason, 1975;
Correspondence: sity of Miinster,
W. Hubert, Institute Rosenstr. 9, D-4400
of Psychology I, UniverMiinster, F.R.G.
Lang, 1979; Lang et al., 1980). There is considerable evidence for the latter view indicating that autonomic activity is differentiated along multiple evaluative dimensions of emotion (e.g., Ax, 1953; Ekman et al., 1983; Schwartz et al., 1981; Winton et al., 1984; Lang et al., 1980) although the patterns of autonomic activity associated with different emotions have not yet been clearly identified. Changes in heart rate (HR) are innervated by both the sympathetic and parasympathetic systems, and it is assumed that HR is physiologically and psychologically determined (Lacey et al., 1963; Obrist, 1976, Fowles, 1980). Electrodermal activity (EDA), on the other hand, is innervated entirely by the sympathetic system and may unambiguously reflect primary arousal. Future research is required to investigate changes in HR and EDA and their potential interaction to distinct emotional stimuli.
132
Since it had been difficult to demonstrate a clear relationship between specific physiological processes and specific emotions, the question arised; how bodily sensations relate to physiological changes and both to subjective emotional experiences. Some studies demonstrated a differential perception of bodily sensations in association with different recalled emotional experiences (e.g., Fahrenberg, 1965; Landy and Stern, 1971; Shields, 1984; Scherer and Tannenbaum, 1986). But acute changes in perceived bodily sensations were rarely assessed (e.g., Tyrer et al., 1978; de Jong-Meyer et al., 1990). All three relevant response components (physiological recordings, self-report of mood and perceived bodily sensations) were investigated in a few studies only (e.g., Pennebaker, 1982; de JongMeyer et al., 1990). From these studies it can be concluded, that relationships between objectively recorded physiological changes and perceived changes show a very high interindividual variability and are generally low. Relationships between bodily sensations and affective states, on the other hand, are more consistent and show some specificity, particularly regarding unpleasant valenced states like anxiety, sadness or anger. Several findings support the conclusion that emotional experiences are accompanied by changes in the endocrine system (for review see Brady, 1975; Mason, 1975; Henry, 1986). Bermond and Frijda (1987) have argued that endocrine feedback mechanisms may regulate the intensity of emotional experiences. Emotion-induced changes in the level of hormones, often a relatively slow process, may not have direct effects on the emotional experience, but may predispose the individual to react in a certain way during future emotion eliciting situations. Whereas important in endocrine stress research (see review by Rose, 1984) the pituitary-adrenalcortical system has rarely been investigated in the area of differential emotion research. More generally, anxiety (e.g., Bloch and Brackenridge, 1972; Brady, 1975; Brown and Heninger, 1975; Francis, 1979; Hubert and de Jong-Meyer, 1989; Hubert et al., 1989) depression (e.g., Sachar, 1981) as well as bereavement and sadness (e.g., Hofer et al., 1972a, b; Jacobs et al., 1985; Irwin et al., 1988) have been associated with an elevation of cortisol, while re-
laxation and meditation on the other hand were found to correlate with declines in cortisol levels compared to control conditions (e.g., Jevning et al., 1978, DeGood and Redgate, 1982). Systematic research on positive emotions, for example joy or happiness, are lacking. Research on endocrine response patterns during and following psychological stimuli has often been restricted by the collection of blood samples. The possibility of assessing hormones in saliva is of great advantage, since saliva samples are easily and repeatedly obtainable and do not demand stressful venipuncture (see review by Riad-Fahmy et al., 1982). Fortunately, saliva flow-rate does not influence salivary cortisol levels. It could be demonstrated that there is an excellent correlation (r > 0.90) between serum free and saliva cortisol levels. Therefore, salivary cortisol measurements would appear to be the method of choice in human psychoendocrinological research (Bassett et al., 1987). Film stimuli are often used as elicitors of mood changes. McHugo et al. (1982) for example, demonstrated that film segments can reliably elicit both pleasant and unpleasant valenced affective changes. Across films a wide range of affective responses can be induced, thus film segments represent one rich source of stimuli for emotional research. The present film study was designed to induce changes in affective states, which were assumed to be accompanied by differential response patterns in perceived bodily sensations, autonomic activity and salivary cortisol. To get differential changes in affect, film content was chosen to evoke a pleasant, amusing versus an unpleasant, arousing state of suspense. Suspense, not unpleasantness per se, was shown in prior film studies to elicit cortisol changes (Hellhammer et al., 1986). The other assumptions, based on previous film and other mood induction results were, that perceived bodily changes would be enhanced after the suspense scenes, that electrodermal activity would be increased during the suspense and decreased during the pleasant, amusing film because of the different arousal impact, and that the HR deceleration (because of the shared information intake type of situation, Lacey et al., 1963) would be reduced
133
during the suspense film by stimuli high in preparedness (Cook et al., 1986) potential for avoidance tendencies (snakes, spiders).
MATERIALS
AND
METHODS
Subjects The subjects were 20 healthy male students (excluding psychology students) aged 21-24 years (M = 22.7 years). Written informed consent was obtained from all. They were paid DM 50 for participation in the experiment. Subjective rating scales Mood ratings were assessed using eight visual analogue scales (VAS; Aitken, 1969; see Fig. 1 for content). A list of bodily sensations consisted of 14 VAS items (see Fig. 2 for content). For each item the subject had to state on a lOO-mm line (0: not at all, 100: very much) to what extent he experienced the corresponding mood/bodily sensation at the moment. These items were shown in previous studies to discriminate between different affective states (Hubert and de Jong-Meyer, 1989; de Jong-Meyer et al., 1990). Apparatus HR (bpm) was recorded by conventional ECG (lead II) using Ag/AgCl electrodes filled with ECG paste. ECG signal was amplified by Natic UBV300 and the R-R intervals were scored and converted to HR using a Kontron PSI980 computer. EDA was recorded by a constant voltage of 0.5 V with Beckman Ag/AgCl electrodes (8 mm diameter) filled with 0.05 M NaCl Unibase electrode paste. The recording was taken from thenar-hypothenar sites of the left hand using a Natic EDA coupler. The original DC signal (skin conductance level, SCL, pS) and the AC signal with a time constant of 10 s were sampled at 10 Hz and recorded on disk. Criterion for spontaneous fluctuations (SFs, rate/mm) in skin conductance was 2 0.1 pS. Salivary cortisol Subjects collected l-2 ml saliva at -15, 0, 15, 30 and 45 min referring to the onset of the film.
Saliva samples were stored at - 20” C until assayed. Salivary cortisol was determined in duplicate by using a commercial radioimmunoassay kit from Mallinckrodt Diagnostica (SPAC Cortisol, Dietzenbach, F.R.G.). The intra- and interassay coefficients of variation were 1.9-2.3% and 8.4%, respectively. Procedure Two lo-min film segments were used as emotional stimuli. They were presented in counterbalanced order on consecutive days (one film per day). The suspense segment from ‘Indiana Jones Raiders of the lost Ark’ (U.S.A. 1980, directed by S. Spielberg) showing also situations with snakes and spiders were chosen as the unpleasant, arousing stimulus. Scenes from ‘Peanuts’ (TV program) were chosen as the pleasant and amusing stimulus. The film segments had been rated by a different sample of subjects as eliciting a calm, joyful or an unpleasant, arousing mood, respectively. In an introductory session (1 h) subjects got familiarized with the procedure and equipment. The two experimental sessions (1.5 h) were conducted in a light-controlled and soundproof laboratory, with the subject seated in a comfortable semi-reclining chair in front of a TV monitor. Subjects were instructed to refrain from alcohol or other drugs and exercise prior to the sessions. To control for circadian variations of salivary cortisol all sessions were conducted between 13.00 and 14.30 or between 15.00 and 16.30 h (randomly assigned). Sessions followed the same schedule: (a) attachment of electrodes, (b) 3-min baseline recording, (c) lo-min film presentation, (d) 35-min post film period. Onset of the film was 30 min after arrival. Physiological recordings were done continuously during baseline and film presentation. Mood and bodily sensations were recorded before and following film presentation. Data scoring and analysis The values of HR and SCL were reduced to one average for 60-s periods. For SFs time period was also 60-s. For the physiological recordings, mean values of the last 2 min of the baseline period were used as baseline scores. For cortisol the baseline levels were at time 0-min.
134
For analyses of variance (ANOVA) and multivariate analyses of variance (MANOVA), Greenhouse-Geisser corrections on the degrees of freedom were applied where appropriate.
Mood Item
Tired
RESULTS
Relaxed
I
_I
-
1
i-r--
-
~-~
~ Mood Differences among baselines of all mood items for the two films were not significant. A 2 (Film) x 2 (Time) repeated-measures MANOVA was performed for all mood items as dependent variables. It revealed a significant Time main effect (F&11) = 4.17, P = 0.016) and a significant Film x Time interaction (F&11) = 8.40. P < 0.001). The main effect for Film was not significant. Follow-up univariate ANOVAs of the significant multivariate Time main effect and multivariate Film x Time interaction produced significant Time main effects for ‘relaxed’ (F(1,9) = 22.76, P = 0.001) and ‘curious’ (F(1,9) = 19.22, P = 0.002), while the Film x Time interaction was significant for ‘relaxed’ (F(l,S) = 8.02, P = 0.022), ‘joyful’ (F(1,8) = 22.09, P = 0.002) and ‘irritated’ (F(1,8) = 16.00, P = 0.004). As illustrated in Fig. 1, there were increases for ‘irritated’ (F(1,6) = 9.66, P = 0.021), and decreases for ‘relaxed’ (F(1,6) = 24.44, P = 0.003) and ‘joyful’ (F(1,6) = 11.14, P = 0.015) during the suspense scenes. On the other hand, the pleasant, amusing film segment induced an increase for ‘joyful’ (F(1,9) = 10.94, P = 0.009). There was a decline for ‘curious’ after both films (negative: F(1,9) = 21.80, P = 0.001; positive: F(1,9) = 7.93, P = 0.020). Bodily sensations Differences among baselines of all bodily sensation items for the two films were not significant. A 2 (Film) x 2 (Time) repeated-measures MANOVA was performed for all bodily sensation items as dependent variables. It revealed a significant Film X Time interaction (F(I4,5) = 5.51, P = 0.035). The main effects for Film or Time were not significant.
Joyful
Sad
Irritated
Curious
Vigorous
Anxious
i
i_ ~.
-1
-30
-20
-A----10
0
~~_~ 10
20
30
VAS-Diff. m Fig. 1. Mean
pleasant/amuwng
r=-i unpleasant/arousing
changes in mood during pleasant, unpleasant, arousing film stimulation.
amusing
and
Follow-up univariate ANOVAs of the significant multivariate Film X Time interaction produced significant Film X Time interactions for ‘hot face’ (F(1,4) = 8.56, P = 0.043), ‘sweating’ (F(1,4) = 9.26, P = 0.038), ‘heart rate increasing’ (F(1,4) = 17.94, P = 0.013), ‘sweaty palms’ (F (1,4) = 28.39, P = 0.006), ‘restless’ (F(1,4) = 10.87, P = 0.030) and ‘difficulties in breathing’ (F(1,4) = 12.58, P = 0.024). As illustrated in Fig. 2, there were significant increases for ‘hot face’ (F(1,4) = 10.25, P = 0.033), ‘sweating’ (F(1,4) = 15.18, P = 0.017), ‘heart rate increasing’ (F(1,4) = 23.37, P = 0.008), ‘sweaty palms’ (F(1,4) = 32.68, P = O.OOS), ‘restless’ (F(1,4) = 20.26, P = 0.010) and ‘difficulties in breathing’ (F(1,4) = 11.46, P = 0.027) during the suspense scenes, while during the pleasant, amusing film ‘restless’ (F(1,4) = 7.76, P = 0.049) increased significantly.
135
11.70, P < 0.001) and SFs (F(6,116) = 6.69, P < 0.001). Multiple t-tests were performed to compare mean changes during film stimulation to baseline levels. HR decreased from minutes 2 to 4 (for all P -Z 0.05) during the suspense scenes and from minutes 2 to 5 (for all P < 0.05) during the pleasant, amusing film segment (Fig. 3). For SCL there were significant increases for minutes 3 and 6 (both P < 0.01) during the suspense scenes, and after an initial increase for minute 1 (P < 0.01) there were significant decreases from minutes 5 to 10 (for all P < 0.05) during the amusing film (Fig. 4). SFs increased significantly for minute 1 and for minutes 3 to 10 (for all P < 0.05) during the suspense film and decreased insignificantly after an initial increase for minute 1 (P < 0.05) for minutes 3 to 8 during the amusing film (Fig. 5).
Bodily sensations item Hot face Feeling
cold
Sweating Shallow
breathing
HR increasing Breathing
faster
Blood rush. to head Tense
atomaoh Trembling
Close Paralyaed
to tears with fear
Sweaty
palm8
Ffeatleasnesa Diff.
in breathing
HR
VAS-Diff. m
positive
Fig. 2. Mean changes in the perception during pleasant, amusing and unpleasant, lation.
Autonomic
negative of bodily sensations arousing film stimu-
activity
Differences among baselines for the two films were not significant. A 2 (Film) x 11 (Time) repeated-measures multivariate analysis of variance (MANOVA) was performed for HR, SCL, and SFs as dependent variables. It revealed significant main effects for P = 0.016) and Time Film (1;(1,6) = 10.90, (F(6,108) = 4.94, P < 0.001). The Film X Time interaction was also significant (F(5,85) = 5.25, P < 0.001). Follow-up univariate ANOVAs produced significant main effects for Film on the measures of SCL (F(1,19) = 9.58, P = 0.006) and SFs (F(1,19) = 34.14, P < O.OOl), and for Time on the measures of HR (F(4,74) = 7.07, P = 0.002), SCL (F(2,46) = 4.04, P = 0.024) and SFs (F(6,116) = 3.79, P < 0.001). Furthermore, the Film X Time interaction was significant for SCL (P(2,46) =
73
71 c
69t
67
1 7%
66’
-’
0
+
pleasant/amusing
’ 1
’ 2
3 Time
Fig. 3. Mean changes amusing and unpleasant,
4
unpleasant/arousing
,
1
5
6
in 60-s
1-d 7
6
9
IO
intervalls
in heart rate (HR) during pleasant, arousing film stimulation (0 indicates baseline).
136
Salivury cortisol Differences among baselines for the two films were not significant. A 2 (Film) X 5 (Time) repeated-measures ANOVA for salivary cortisol levels revealed a significant main effect for Time (F(2,30) = 56.96, P < 0.001). The main effect for Film and the Film X Time interaction were not significant. As illustrated in Fig. 6 salivary cortisol levels declined during both conditions.
l/min 61
DISCUSSION The aim of this study was to use multimodal assessment of subjective, autonomic and endocrine variables to investigate differential responding to emotion-inducing film segments. 1
’ +
pleasantlamuslng
0
SCL
0
1
2
3
++
1
1
4
5
unpleasant/arousing
I 6
7
8
9
IO
Time in 60-s intervalls Fig. 5. Mean changes in spontaneous fluctuations of skin conductance (SFs) during pleasant. amusing and unpleasant, arousing film stimulation (0 indicates baseline).
te-
+ 0
if
pleasant/amusing
1
2
3
4
unpleasant/arousing
1
1
1
1
5
6
7
8
‘~.
9
10
Time in 60-s intervalls Fig. 4. Mean responses in skin conductance level (SCL) during pleasant, amusing and unpleasant, arousing film stimulation (0 indicates baseline).
The results showed that film segments induced an unpleasant, irritated and arousing versus joyful state. Mood profiles differed clearly and thus showed, that film stimuli can be used for differentiated mood induction purposes, and are in line with similar findings (e.g., Lazarus et al. 1962; Tyrer et al., 1978; McHugo et al., 1982; Hubert and de Jong-Meyer, 1989). The evoked changes in perceived bodily sensations supported our assumptions. After the suspense scenes subjects reported changes in temperature related sensations (hot face, sweating, sweaty palms), difficulties in breathing, an increased HR and more restlessness, while the amusing film elicited a small increase in restlessness and no changes in other bodily sensations. The results are in line with Tyrer et al. (1978), Pennebaker (1982) and de Jong-Meyer et al. (1990). Furthermore, these findings expand results of Fahrenberg (1965).
137
Salivary
cortisol
18
16, ‘9
14.
12
87 + 6 I I___~~ -15
pleasant/amusing
+
unpleasantlarousfng
1
I
0
15
Time
30
45
[mid
Fig. 6. Mean changes in salivary cortisol before and following pleasant, amusing and unpleasant, arousing film stimulation.
Landy and Stern (1971), Shields (1984) and Scherer and Tannenbaum (1986), who demonstrated differential perceptions of bodily sensations in association with different recalled emotional experiences. The results point to the importance of assessing perceived bodily sensations as a component of emotional responding, particularly, if one considers the only mild to moderate range of reported mood changes. Future studies should include the assessment of bodily sensations to decide, whether other affective states are accompanied by distinct profiles of these sensations, and whether they are indicative of specific unpleasant emotional states or of different bodily arousal patterns. The finding, that pleasant and/or relaxed states are characterized by a lack of changes in bodily sensations, if replicated, has important implications for the elaboration of treatment strategies for psychosomatic or affective disorders.
On the autonomic level there were differential response patterns for SCL and SFs. The significant increase during the first minute of both films may be interpreted as an orienting response. The increases in SCL and SFs during the suspense scenes are similar to results of Lazarus et al. (1962) Mordkoff (1964) Kaiser and Roessler (1970) and Steptoe and Vbgele (1986) using also unpleasant and arousing film segments. The decrease in SCL and SFs during the amusing scenes were found in a study of Kaiser and Roessler (1970) too. Regarding HR, we found decelerations for both film segments and no differences between them. This is in line with the suggestion of Lacey et al. (1963) that situations with the intake of informations are characterized by HR decelerations. We had assumed, however, that because of the high potential of our suspense film to provoke avoidance tendencies, HR deceleration would be reduced during this film, leading to differences compared to the amusing one. Although the relevant Film X Time interaction failed to become significant, the results do not argue strongly against this assumption. Subjectively, the participants got more irritated, and there were stronger HR accelerations during the second part of the suspense segment, while during the amusing film this trend was weaker towards the end. The literature regarding the effects of film stimuli on HR is mixed. Whereas Lazarus et al. (1962) Mordkoff (1964) and Steptoe and Vbgele (1986) reported significant increases of HR during the presentation of unpleasant, arousing film stimuli, Carruthers and Taggart (1973) demonstrated a HR deceleration as a response to a violent movie, even in the presence of increased sympathetic activity. Van den Akker and Steptoe (1987) could not demonstrate an acceleration in HR during an industrial accident film. Further studies should test the functional significance of HR changes in relation to the information intake versus rejection assumption. They should allow for more objective control of this information processing dimension and then vary the valence of the elicitor. The expected increase in salivary cortisol levels in response to suspense could not be observed.
13X
During both conditions there were decreases in salivary cortisol levels. Cortisol in serum and saliva has a marked circadian rhythm (Riad-Fahmy et al., 1982). During morning and afternoon hours cortisol levels decrease continuously. This circadian rhythm seems to be the reason for the decrease in salivary cortisol levels in this study. One may argue that the time course of saliva sampling is not sufficient to detect increases in salivary cortisol levels. But Riad-Fahmy et al. (1983) demonstrated that time intervals of 15 min provide a convenient way to estimate salivary cortisol variations. Another reason for the lacking activation of the adrenal system during suspense might be the short duration of the film and/or its mild to moderate affective changes. Other studies demonstrating a significant cortisol elevation used longer films of 30 to 90 min duration (e.g., Bernick et al., 1971; Brown and Heninger, 1975; Hellhammer et al., 1986; Bossert et al., 1988; Hubert and de Jong-Meyer, 1989) and stress of venipuncture, a was shown to be effective in strong stimulus, inducing cortisol elevations (Mason et al.. 1973; Rose and Hurst, 1975; Hubert et al., 1989). Salivary cortisol declined more rapidly prior film viewing compared to the period during and following film presentation. One may argue that the setting up procedures for the experiment were stressful and induced a cortisol release with a subsequent rapid decrease. But inspection of the mood and bodily sensation ratings at baseline do not support this argumentation. One explanation for this initial rapid decrease might be that physical exercise (e.g., Cook et al.. 1987; O’Connor and Corrigan, 1987) produced an elevation in cortisol level while subjects came to the laboratory. The first saliva sample was assessed already 15 min after arrival. Future studies should extent the time of adaptation in order to minimise such effects. Taken together, our results as well as those from other film studies prove this paradigm to be useful in eliciting mild to moderate affective states and their concomitants. Further investigations should include more different emotions, particularly distinct pleasant and unpleasant ones, varying in degree of arousal, and should control attentional set and direction of information processing. Differential perceived bodily sensations as well as
the contribution of HR changes to differential emotion profiles seem to be promising. Furthermore, psychoendocrinological investigations should use longer film segments to study cortisol response patterns during different emotions. Given that subsequent studies with a larger sample of males and females replicate and extend these findings we may learn more on pattern classification of different emotions.
ACKNOWLEDGMENTS This study was supported partly by a grant Forschungsgemeinschaft (Jo from the Deutsche 140/l -2). The authors thank Prof. E. Nieschlag of Reproductive and Mrs. M. Miiller (Institute University of Miinster) for cortisol Medicine, analyses.
REFERENCES Altken. R.C. (1969) A growing edge of measurement of feelings. Proc. Roy. SW. Med.. 62: 989-993. Ax. A.F. (1953) The physiological differentiation between fear and anger in humans. Pq&xon~. Med., 15: 433-442. Bassett, J.R., Marshall, P.M. and Spillane, R. (1987) The physiological measurement of acute stress (public speaking) in bank employees. Inr. J. Psychoph~vsiol., 5: 265-213. Bermond, B. and Frijda, N.H. (1987) The intensity of emotional feelings: Product of peripheral emotional responses or cognitions? Commun. Cogn., 20: 191-205. Bernick. N., Kling, A. and Borowitz, G. (1971) Physiologic differentiation of sexual arousal and anxiety. P.~ychmom. Med.,
33: 341-352.
Bossert, S., Berger, M.. Krieg, J, C., Schreiber, W.. Junker, M. and von Zerssen, D. (1988) Cortisol response to various stressful situations: Relationship to personality variables and coping styles. Neuropsychohiology. 20: 36-42. Brady, J.V. (1975) Toward a behavioral biology of emotion. In L. Levi, (Ed.), Emorron.s Their prrramerer.~ and memmment. Raven Press. New York, pp. 17-45. Brown, W.A. and Heninger, G. (1975) Cortisol. growth hormone. free fatty acids. and experimentally evoked affective arousal. Am. J. P.ychmr.. 132: 1172-1176. Cannon, W.B. (1927) The James-Lange theory of emotion: A critical examination and an alternative theory. Ant. J. Psychol., 39, 106-124. Carruthers, M. and Taggart, P. (1973) Vagotonicity of violence: Biochemical and cardiac responses to violent film5 and television programmea. Bnr. Med J., 18: 384-389.
139 Cook III, E.W., Hodes, R.L. and Lang, P.J. (1986) Preparedness and phobia: Effects of stimulus content on human visceral conditioning. J. Ahn. Psychoi., 95: 195-207. Cook. N.J., Ng, A., Read, G.F., Harris, B. and Riad-Fahmy, D. (1987). Salivary cortisol for monitoring adrenal activity during marathon runs. Harm. Res., 25: 18-23. DeGood, D.E. and Redgate, E.S. (1982) Interrelationship of plasma cortisol and other activation indices during EMG biofeedback training. .I. Behau. Med., 5: 213-223. De Jong-Meyer. R., Hubert, W., KGrner, U. and Geyer. E. (1990) Induktion von Angst and Freude fiber die Imagination autobiographischer Erinnerungen [The induction of anxiety and happiness by the imagery of autobiographical recollection]. 2. Klin. Psychol., 19: 83-95. Ekman, P.. Levenson, R.W. and Friesen, W.V. (1983) Autonomic nervous system activity distinguishes among emotions. Science, 221: 1208-1210. Fahrenberg, J. (1965) Zur Frage einer differentiellen Physiologie der Affekte [Questions about a differential physiology of affects]. Psycho. Forschung, 28: 422-438. Fowles. D.C. (1980) The three arousal model: Implications of Gray’s two-factor learning theory for heart rate, electrodermal activity, and psychopathy. Prychophysmlog);. 17: X7-104. Francis. K.T. (1979) Psychologic correlates of serum indicators of stress in man: A longitudinal study. Ps)?chosom. Med., 41: 617-628. Hellhammer, D.H., Rottger, K.. Lorenzen, J. and Hubert, W. (1986) Suspense increases salivary cortisol. In R.C. Fredrickson, H.C. Hendrie. J.N. Hingtgen and M.H. Aprison (Eds.), Neuroregulation of au~onomrc, endocrine and immune systems, Martinus Nijhoff, Boston, pp. 521-524. Henry, J.P. (1986) Neuroendocrine patterns of emotional response. In R. Plutchik and K. Kellerman (Eds.), Emorion Theory, research, and experience. Vol. 3, Biological foundations of emotion, Academic Press, Orlando, pp. 37-60. Hofer. M.A., Wolff, C.T., Friedman, S.B. and Mason, J.W. (1972a) A psychoendocrine study of bereavement. Part I. 17-hydroxycorticosteroid excretion rates of parents following death of their children from leukemia. Psychosom. Med.. 34: 481-491. Hofer. M.A., Wolff, C.T., Friedman, S.B. and Mason, J.W. (1972b) A psychoendocrine study of bereavement, Part Il. Observations on the process of mourning in relation to adrenocortical function. Psychosom. Med., 34: 492-504. Hubert, W. and de Jong-Meyer, R. (1989) Emotional stress and saliva cortisol response. J. Clin. Chem. C/in. Biochem., 27: 235-237. Hubert, W., Miiller, M. and Nieschlag, E. (1989) Stress reactions in response to the procedure of LHRH-tests as measured by salivary and serum cortisol and psychological variables. Harm. Res., 32: 198-202. Irwin. M., Daniels, M., Risch, SC., Bloom, E. and Weiner, H. (1988) Plasma cortisol and natural killer cell activity during bereavement. Biol. Psychiat.. 24: 173-178. Jacobs. S.. Mason, J., Kosten, T., Kasl, S., Ostfeld, A.. Atkins,
S., Gardner, C. and Schreiber, S. (1985) Acute bereavement, threatened loss, ego defenses and adrenocortical function. Psychother. Psychosom.. 44: 151-159. James, W. (1890) The principles OJ psychology, Henry Holt, New York. Jevning, R., Wilson, A.F. and Davidson, J.M. (1978) Adrenocortical activity during meditation. Harm. Behau., 10; 54-60. Kaiser, C. and Roessler, R. (1970) Galvanic skin responses to motion pictures. Pert. Mot Sk.. 30: 371-374. Lacey, J.I., Kagan, J., Lacey, B.C. and Moss, H.A. (1963) The visceral level: Situational determinants and behavioral correlates of autonomic response patterns. In P. Knapp (Ed.), Expressions of the emotions in man. International University Press, New York. pp. 161-196. Landy. F.J. and Stern, R.M. (1971) Factor analysis of a somatic perception questionnaire. J. Psychosom. Res.. 15: 179-181. Lang, P.J. (1979) A bio-information theory of emotional imagery. Psychoph_v.volo~. 16: 495-512. Lang, P.J., Kozak, M.J., Miller, G.A.. Levin, D.N. and McLean, A. (1980) Emotional imagery: Conceptual structure and pattern of somato-visceral response. Psychophysrology, 17: 179-192. Lazarus. R.S., Speisman, J.C., Mordkoff, A.M. and Davison, L.A. (1962) A laboratory study of psychological stress produced by a motion picture film. Psychol. Monogr., 76: Whole No. 553. Mason, J.W. (1975) Emotion as reflected in patterns of endocrine integration. In L. Levi (Ed.), Emotions - Their parameters and measurement, Raven Press. New York, pp. 1433 181. Mason. J.W., Hartley, L.H.. Kotchen, T.A., Mougey, E.H., Ricketts, P.T. and Jones, L.G. (1973) Plasma cortisol and norepinephrine responses m anticipation of muscular exercise. Psychosom. Med.. 35: 406-414. McHugo, G.J.. Smith, C.A. and Lanzetta. J.T. (1982) The structure of self-reports of emotional responses to film segments. Motiu. Emot.. 6: 365-385. Mordkoff. A.M. (1964) The relationship between psychological and physiological response to stress. Psychosom. Med.. 26: 135-150. Obrist, P.A. (1976) The cardiovascular-behavioral interaction As it appears today. Psychophysiology, 13: 95-107. O’Connor, P.J. and Corrigan. D.L. (1987) Influence of shortterm cycling on salivary cortisol levels. Med. SCL Sports Exert., 19: 224-228. Pennebaker, J.W. (1982) The psychology of physical symptoms. Springer-Verlag, New York. Riad-Fahmy, D., Read. G.F. and Walker, R.F. (1982) Steroids in saliva for assessing endocrine function. Endocr. Rev.. 3: 367-395. Riad-Fahmy, D., Read, G.F. and Walker. R.F. (1983) Salivary steroid assays for assessing variation in endocrine activity. J. Steroid Biochem., 19: 265-272. Rose, R.M. (1984) Overview of endocrinology of stress. In G.M. Brown, (Ed.). Neuroendocrmology cmdpsychmtric disorder, Raven Press. New York. pp. 95-122.
140 Rose. R.M. and Hurst. M.W. (1975) Plasma cortisol and growth hormone responses to intravenous catheterization. J. Human Stress, 1: 22-36. Sachar, E.J. (1981) Psychobiology of affective disorders. In E.R. Kandel and J.H. Schwartz (Eds.), Principles of neural .scrence, Elsevier. New York, pp. 61 I-61 9. Schachter, S. and Singer, J. (1962) Cognitive, social, and physiological determinants of emotional state. Psycho/. Ret:.. 69, 379-399. Scherer, K.R. and Tannenbaum. P.H. (1986) Emotional experiences in everyday life: A survey approach. Mob. Emot., 10: 295-314. Schwartz. G.E.. Weinberger. D.A. and Singer, J.A. (1981) Cardiovascular differentiation of happiness. sadness, anger. and fear following imagery and exercise. P.rycho.wm. Med.. 43: 343-364.
Shields, S.A. (1984) Reports of bodily change in anxiety, sadness, and anger. Motic,. Emot., 8: l-20. Steptoe, A. and VGgele, C. (1986) Are stress responses influenced by cognitive appraisal? An experimental comparison of coping strategies. Brrt. J. PsychoI.. 71: 243-255. Tyrer. P.. Lewis. P. and Lee. I. (1978) Effects of subliminal and supraliminal stress on symptoms of anxiety. J. Nem. Ment. DlS., 166: 88-95. Van den Akker, 0. and Steptoe, A. (1987) Psychophysiological responses in women with premenstrual and menstrual symptoms. J. Psychoph_ysiol., 1: 149-158. Winton. W.M., Putnam. L.E. and Krauss, R.M. (1984) Facial and autonomic manifestations of the dimensional structure of emotion. J. Exp. Sot. Ps_vchol.. 20: 195-216.