Psychophysiological correlates of affects in human olfactionCorrelats neurophysiologiques des états affectifs déclenchés par les odeurs chez l'homme

https://doi.org/10.1016/S0987-7053(02)00339-8Get rights and content

Abstract

Pleasantness and arousal have been identified as the main dimensions of affective responses to environment. Pleasantness is defined as the degree of favorable feelings a subject can experience under given circumstances. Arousal is defined as the degree of excitement (general activation) the subject feels under these circumstances. In visual and auditory modalities, many studies using measures such as facial electromyographic (EMG) activity and skin conductance (SC) have found those parameters to vary as a function of either pleasantness or arousal: for example, facial corrugator EMG activity covaries with the pleasantness dimension, while SC increases together with arousal. The first objective of this research is to study the possible covariation between peripheral measures and pleasantness/arousal in olfaction. We also examined the effect of odor intensity on facial and autonomic variations. The second objective was to investigate whether odors could evoke verbally specific emotions (e.g. joy, anger, fear, surprise, disgust or sadness) and also induce specific patterns of peripheral responses. Participants were exposed to 12 different odors while their facial and autonomic parameters were recorded, and estimated their intensity, pleasantness, and arousal capacity. Then, they chose between seven words for emotions (fear, anger, sadness, surprise, neutral, joy or disgust) to describe their reaction to the odor. As in vision, olfactory pleasantness covaries (negatively) with facial activity of the corrugator muscle, and arousal (positively) with skin conductance. No relationships were observed between physiological changes and variations in perceived intensity. Results also showed that emotions of “disgust” and “joy” were more frequently evoked verbally than any other emotions, and that only facial EMG activity distinguishes them (e.g. “disgust” vs. “joy” and “neutral state”). The results are discussed in terms of possible existence of two brain systems (defensive and appetitive), each of them being able to vary in metabolic arousal.

Résumé

Les dimensions principales de la réponse affective à l’environnement ont été définies comme le plaisir et l’éveil émotionnel. Ils se définissent respectivement comme le niveau de sensations favorables que le sujet éprouve envers une situation, et le niveau d’excitation que le sujet ressent de cette situation. Nombre de travaux portant sur les modalités auditive et visuelle ont montré par des mesures comme celle de l’électromyogramme (EMG) facial ou de la conductance de la peau (CP), que ces paramètres varient en fonction soit du plaisir, soit de l’éveil : par exemple l’activité du muscle facial Corrugator covarie négativement avec la dimension de plaisir, alors que la CP augmente de façon linéaire avec l’éveil. Le premier objectif de ce travail était d’étudier, en olfaction, la covariation entre plaisir et éveil et les mesures périphériques. Nous avons aussi examiné l’effet de l’intensité de l’odeur sur les changements faciaux et végétatifs. Le deuxième objectif était de savoir si les odeurs peuvent évoquer des associations verbales à des émotions spécifiques (joie, colère, peur, surprise, dégoût ou tristesse) et aussi induire des patrons spécifiques de réponses périphériques. Les participants ont été exposés à 12 odeurs différentes, alors qu’on enregistrait l’activité faciale et végétative. Ils ont ensuite évalué le plaisir, l’éveil et l’intensité. Les sujets devaient enfin choisir parmi sept noms d’émotion (peur, colère, tristesse, surprise, neutre, joie ou dégoût) celui qui décrivait le mieux leur réaction à l’odeur. Les résultats montrent que dans la modalité olfactive comme en vision, le plaisir varie en relation inverse avec l’activité faciale du muscle corrugateur, et l’éveil émotionnel avec l’amplitude de la CP. On n’a pas montré de relation entre ces changements physiologiques et l’intensité perçue. L’étude montre également que les émotions de “dégoût” et de “joie” sont mentionnées verbalement plus que les autres émotions, et que seule l’activité EMG faciale les discrimine les unes des autres (exemple : « dégoût » vs. « joie » et « neutre »). Les résultats sont discutés en relation avec l’existence possible de deux systèmes cérébraux (défensif et appétitif) pouvant varier chacun en activation métabolique.

Introduction

Arousal and pleasantness have been identified as the main dimensions of affective responses to the environment 〚21〛. Pleasantness is defined as the degree of favorable feelings a subject can feel under given circumstances, while arousal is defined as the degree of excitement the subject feels.

In order to provoke emotions in a laboratory setting, investigators have assembled sets of pictures 〚18〛, sounds 〚5〛 and words 〚4〛 especially chosen to give rise to a wide range of emotions: from pleasure to displeasure via neutral emotions. Results indicate that the shape of distributions is very similar in visual as in auditory modality, with arousal ratings increasing with emotional valence (either pleasantness or unpleasantness). According to these authors, results are consistent with the hypothesis that emotion stems from two underlying neural systems: appetitive (for positive affective states) and defensive (for negative affective states), and that each varies in arousal (see 〚8〛 for a review). Arousal is defined here as the intensity of the activation of these subsystems, and not as a separate subsystem.

Besides such subjective evaluation, it is possible to study objectively the activation of these systems by examining peripheral and central responses to emotional stimuli 〚19〛. Many studies suggest that dimensions of pleasantness and arousal covary consistently with other measures of reactivity in physiological systems. Lang and his colleagues 〚6〛, 〚15〛, 〚19〛 have recently generated a large database indicating that visceral responses, as well as facial displays of emotion, systematically vary with differences in pleasantness or arousal. For example, reports of pleasantness covary systematically with heart rate activity, facial responses and startle reflex, while reports of arousal covary systematically with skin conductance variations and late event-related potentials. Thus, different physiological systems are tuned to different dimensions. The present study was designed in order to examine whether these relationships between pleasantness and arousal on the one hand and physiological systems on the other hand could be observed when olfactory stimuli were perceived.

Pleasantness dimension of odors is so prominent that authors have used odorants to induce emotional states (〚13〛, 〚26〛 for a review). Reports from patients with persistent olfactory impairment (e.g. partial or general anosmia) also underscore the importance of odor pleasantness by suggesting that there is a lower quality of life related to the loss of the ability to detect pleasant odors 〚22〛, 〚31〛. Compared to the pleasantness dimension, little is known about the arousal dimension in olfaction.

In addition to odor pleasantness, investigations using multidimensional methods suggested that odor intensity has also to be taken into account in olfaction studies, especially when human subjects are asked to categorize odorants 〚14〛, 〚27〛.

Studies in the olfactory modality demonstrated that skin conductance could be modulated by the perception of an olfactory stimulation 〚25〛, 〚32〛. Skin conductance variations were found to be associated with odor concentration 〚30〛. Other studies indicated that electrodermal response variations could be modulated by odor pleasantness 〚2〛, 〚9〛, but dimensions such as arousal, and also odor intensity were not analyzed. In some of these studies 〚1〛, 〚33〛, authors examined the possible link between verbally evoked basic emotions (i.e. anger, fear, sadness, surprise, disgust and happiness) by odors and autonomic nervous system responses. The main findings were that: (1) this link exists when pleasant emotions are considered, and (2) odorants evoked more certain negative basic emotions (disgust in this case) in the “verbal channel” (e.g. subjective reports), while the “visceral channel” (analysis of autonomic nervous system responses) revealed more other kind of negative emotions (anger in this case). With regard to facial EMG recordings, it has been demonstrated that unpleasant odors provoke greater corrugator muscle response than do pleasant ones 〚16〛. Here also, dimensions of odor intensity and arousal were not taken into account.

Even though well established in the visual modality, the link between self-report ratings (of pleasantness and arousal for example) and peripheral variations (such as facial EMG and skin conductance) remains unclear in the olfactory modality. The first aim of the study was to examine the covariation between dimensions of pleasantness and arousal and facial or electrodermal responses, in the olfactory modality. As we have seen, this problem was previously addressed in the visual and the auditory modalities and the main results of those studies suggested that decreased corrugator muscle activity is associated with increasing stimulus pleasantness and that electrodermal response increases linearly with reported arousal. We also considered assessment of odor intensity, given the importance of this dimension, as we have seen above. To address this first aim, participants were instructed to smell odors, while facial and autonomic parameters were recorded, and had to evaluate them on the arousal, intensity and pleasantness dimensions. The second aim was to address a question regarding relationships between specific emotions evoked verbally by odors and physiological changes. Do odors that could evoke verbally specific emotions (e.g. joy, anger, fear, surprise, disgust or sadness) also cause specific patterns of peripheral responses? To address this question, all subjects were instructed to choose between seven words for an emotion (fear, anger, sadness, surprise, neutral, joy or disgust; see 〚12〛) the one best describing their reaction to the odor.

Section snippets

Subjects

The subjects were 12 healthy undergraduate students (six women and six men; mean age: 21.5 years; standard deviation: 1.73) from the Claude Bernard University of Lyon (France). Three of them were smokers and all were right-handers (Edinburgh Laterality Inventory, 〚23〛). Before the experiment began, the experimenter explicitly asked the subjects whether they had any olfactory problems, and none declared any. All subjects gave informed consent.

Odors

Odors were selected among a large set of odorants

Data analysis

Given the small sample of subjects and odors, we decided to analyze the data using non-parametric statistics. Two kinds of analyses were performed. In order to achieve the first aim, we examined relationships between physiological responses (skin conductance and facial EMG corrugator) and dimensions of pleasantness, arousal and intensity. To this end, data were analyzed by odors, using Kruskal–Wallis one-way analysis of variance (ANOVA). Each physiological measure was analyzed according to

Physiological responses and dimensions of pleasantness, arousal and intensity (first aim)

When physiological variables were analyzed according to pleasantness, Kruskal–Wallis one-way ANOVA indicated no significant difference for SC evoked by unpleasant odors versus SC evoked by pleasant odors (χ = 0.410; P > 0.05; Fig. 1A). However, a significant difference was observed for EMG evoked by unpleasant odors versus EMG evoked by pleasant odors (χ = 8.308; P < 0.005; Fig. 1B), indicating that unpleasant odors evoked greater corrugator facial activity than pleasant ones. The same

Discussion

The aims of the present study were (i) to examine the covariation between dimensions of pleasantness and arousal and facial or electrodermal responses, in the olfactory modality, and (ii) to investigate relationships between the verbal evocation of specific emotions (e.g. fear, anger, sadness, surprise, joy and disgust) in presence of odors, and physiological, involuntary changes.

With regard to the first aim, our study showed that, as in other sensory modalities, facial corrugator EMG activity

Acknowledgments

We would like to thank 3 anonymous reviewers for their clear and constructive critics.

References (33)

  • M.M. Bradley et al.

    International affective digitized sounds (IADS): stimuli, instruction manual and affective ratings. Technical report no B-2

    (1999)
  • M.M. Bradley et al.

    Measuring emotion: behavior, feeling, and physiology

  • M.M. Bradley

    Emotion and motivation

  • P. Brauchli et al.

    Electrocortical and autonomic alteration by administration of a pleasant and an unpleasant odor

    Chem Senses

    (1995)
  • J.T. Cacioppo et al.

    Relationships between attitudes and evaluative space: a critical review with emphasis on the separability of positive and negative substrates

    Psychol Bull

    (1994)
  • M.E. Dawson et al.

    The electrodermal system

  • Cited by (0)

    View full text