- Email: [email protected]
The fish is bad: Negative food odors elicit faster and more accurate reactions than other odors
Biological Psychology 84 (2010) 313–317
Contents lists available at ScienceDirect
Biological Psychology journal homepage: www.elsevier.com/locate/biopsycho
The fish is bad: Negative food odors elicit faster and more accurate reactions than other odors S. Boesveldt a,∗ , J. Frasnelli b , A.R. Gordon a , J.N. Lundström a,c,d a
Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA, United States CERNEC, Université de Montréal, Montréal, QC, Canada Dept. of Psychology, University of Pennsylvania, Philadelphia, PA, United States d Dept. of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden b c
a r t i c l e
i n f o
Article history: Received 4 December 2009 Accepted 7 March 2010 Available online 21 March 2010 Keywords: Reaction time Evolutionary psychology Odors Categories Valence Food
a b s t r a c t Dissociating between ‘good’ or ‘bad’ odors is arguable of crucial value for human survival, since unpleasant odors often signal danger. Therefore, negative odors demand a faster response in order to quickly avoid or move away from negative situations. We know from other sensory systems that this effect is most evident for stimuli from ecologically-relevant categories. In the olfactory system the classification of odors into the food or non-food category is of eminent importance. We therefore aimed to explore the link between odor processing speed and accuracy and odor edibility and valence by assessing response time and detection accuracy. We observed that reaction time and detection accuracy are influenced by both pleasantness and edibility. Specifically, we showed that an unpleasant food odor is detected faster and more accurately than odors of other categories. These results suggest that the olfactory system reacts faster and more accurately to ecologically-relevant stimuli that signal a potential danger. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Discrimination between ‘good’ and ‘bad’ odors is arguably of crucial value for human survival. As is well known, unpleasant odors often signal danger, such as spoiled food or a toxin, and, in some non-human species, they even serve as a warning signal of nearby predators (Dielenberg and McGregor, 2001). Negative odors demand a faster response than neutral or pleasant odors because survival depends more often on an organism’s quick response to signals of negative rather than positive situations. Two behavioral studies support this basic assumption by demonstrating that response times of human subjects to unpleasant odors were significantly shorter than for pleasant odors (Bensafi et al., 2002c; Jacob and Wang, 2006). The same principle has been demonstrated in the visual system. Hansen and Hansen showed that an angry face in a crowd of benign or happy faces was detected faster than a happy or benign face in a crowd of angry faces, suggesting that humans are more attentive to threatening signals (Hansen and Hansen, 1988). These findings were later confirmed and extended (Ohman et al., 2001b). The coupling of emotional activation and efficient capture of attention goes beyond faces, as demonstrated by findings that fear-relevant pictures of snakes and spiders were detected faster than fear-
∗ Corresponding author. Tel.: +1 267 519 4688. E-mail address: [email protected] (S. Boesveldt). 0301-0511/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.biopsycho.2010.03.006
irrelevant pictures, such as flowers and mushrooms (Ohman et al., 2001a). These and similar data have been taken as evidence that natural selection has honed the human ability to identify and react to stimuli important for survival (Mineka and Ohman, 2002; Tooby and Cosmides, 1990). Threatening stimuli, such as indicators of spoiled food or the presence of a snake, are prioritized by the cerebral system to effect faster preparation for “fight or flight” action. In the rodent brain, there appears to be a specific neural circuitry – direct linkage of the perceptual and defense systems via the thalamus and amygdala – to achieve this fast mobilization (for a review, please see LeDoux, 2000). There is evidence of a similar system in the human brain (Ohman et al., 2007), also for olfactory stimuli, since the amygdala is located only one synapse away from the olfactory receptors. Moreover, although it is widely assumed that the human olfactory system, unlike the other senses, is independent of thalamic relay, it was recently demonstrated that the thalamus indeed has a functional role in odor processing (Plailly et al., 2008). The transthalamic network is suggested to be a modulatory target of olfactory attentional processing and may serve as an attentional filter, helping to select only those inputs with behavioral relevance for processing downstream of the orbitofrontal cortex (Plailly et al., 2008). Further evidence from research on human subjects demonstrates that biologically relevant stimuli enjoy faster (Lundstrom et al., 2006a) or more direct neuronal pathways (Lundstrom et al., 2008; Morris et al., 1999), resulting in both decreased processing times and lower reaction thresholds than perceptually similar stimuli with no obvious evolutionary importance.
314
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317
Whether the human brain performs categorical olfactory processing akin to that which has been demonstrated in the visual system is not known. Collectively, food odors seem to comprise a category of particular importance to humans: our ability to identify food odors is superior to our performance with odors of nonfood objects (Boesveldt et al., 2009; Fusari and Ballesteros, 2008). Further evidence that the human olfactory system is capable of categorical discrimination comes from the previously-mentioned studies showing that reactions to negative odors are faster than to positive odors (Bensafi et al., 2002c; Jacob and Wang, 2006). Unfortunately, in both of those studies, the pleasant stimulus was a food-related odor (vanillin or amyl acetate) and the unpleasant stimulus was a non-food-related odor (indole or valeric acid), thereby making a direct comparison between odor categories – either pleasant versus unpleasant or food versus non-food – impossible. The present study sought to explore the link between odor processing speed, edibility and valence. To this end, we measured subjects’ reaction time and detection accuracy with two food odors and two non-food odors. In contrast to previous studies, each edibility category was comprised of one pleasant and one unpleasant odor. Based on what is known for the visual system, we hypothesized that, due to its ecological relevance, negative food odors would be processed faster and with higher accuracy than odors from other categories (i.e. pleasant versus unpleasant, food versus non-food). 2. Materials and methods 2.1. Participants A total of 40 young healthy participants (mean age 25 years, range 18–31 years; 20 women) were recruited via posters on the university campus. Subjects were screened prior to inclusion by means of a self-report survey for numerous nasal and neurological disorders known to affect olfactory function. None of the participating women were pregnant, and all had regular menstrual cycles of normal length. Detailed information regarding the experiment was given and written informed consent was obtained from all subjects prior to testing. All aspects of the study were performed in accordance with the University of Pennsylvania IRB and internal regulation at the Monell Chemical Senses Center. 2.2. Stimulus presentation The odors were presented with a computer-controlled 8-channel olfactometer to assure accurate odor onset and a steep odor rise-time. The premise of the olfactometer is that a valve control unit regulates the state of the olfactometer’s eight solenoid valves, each of which directs a continuous airstream of 4 liters per minute (lpm) either into the olfactometer (un-triggered state) or into an odor reservoir (triggered state). When triggered by the valve control unit, a valve directs the airstream into an odor reservoir, and the odorized headspace is transported to the birhinal nosepiece. One of the eight channels serves as a conduit for the control odor and operates in a manner identical to the other channels. Closure of an odor valve triggers the control valve such that whenever an odor is not being delivered, the control air flow is directed to the nosepiece. In the nosepiece, air carrying an odor or the control air mixes with a continuous, low-flow airstream (0.5 lpm). This continuous flow airstream masks the tactile cues that might otherwise alert the subject to channel-switching (Lundstrom et al., submitted for publication). The temporal characteristics of the odor stimulus delivered with the olfactometer settings used in this study (length of tubing, selected airflow rate, etc.) were measured with a photoionization detector (miniPID 200A, Aurora Scientific Inc., Aurora, ON, Canada). The delay to odor delivery, the lapse between the time at which a computer triggers the solenoid valve to the time at which the delivered odor reaches 90% strength, was approximately 450 ms. This calculation is based on the combined lengths of stimulus onset (400 ms, measured from computer trigger) and 10/90% odor rise-time (53 ms, from 10% to 90% odor strength). The stimulus presentation program E-Prime 2.0 Professional (Psychology Software Tools Inc., Pittsburgh, PA) was used to trigger the olfactometer, present the visual cues, and record subjects’ reaction time. 2.3. Odor stimuli Four odors were chosen on the basis of their edibility (food, non-food) and valence. The two food odorants were natural mixtures chosen to mimic the odor of the food object, and the two non-food odorants were single compounds. The food odors were orange (cold-pressed Californian orange oil, Sigma Aldrich, St. Louis,
Fig. 1. Experimental design of reaction time task and timing.
MO, USA) and fish (Fish flavor oil, Givaudan Inc., Geneva, Switzerland), which were diluted with mineral oil (70%, v/v) and 1,2-propanediol (27%, v/v), respectively. The non-food odors were rose (Phenylethyl alcohol, Sigma Aldrich, St. Louis, MO, USA) and dirty socks (Isovaleric acid, Sigma Aldrich, St. Louis, MO, USA), both of which were diluted with 1,2-propanediol to concentrations of 90% (v/v) and 40% (v/v), respectively. Pilot data indicated that these concentrations were iso-intense. Furthermore, each edibility category contained one pleasant odor (orange, rose) and one unpleasant odor (fish, dirty socks). 2.4. Experimental paradigm Perceptual ratings of pleasantness and intensity were obtained by presenting each odorant (10 ml) in individual 100 ml amber glass bottles devoid of visual markers. Odor pleasantness and familiarity were rated on a visual analogue scale (VAS, 10 cm) ranging from ‘Extremely unpleasant/unfamiliar’ to ‘Extremely pleasant/familiar’, and odor intensity was rated on a labeled magnitude scale (LMS, 10 cm) ranging from ‘No sensation’ to ‘Strongest imaginable’. Subjects were seated in a comfortable chair, fitted with the olfactometer nosepiece and in-ear headphones, oriented towards an adjustable computer monitor set at eye-level (1 m viewing distance). To exclude the possibility that auditory cues might influence subjects’ performances, low volume brown noise was played through the headphones throughout the task. To limit the influence of nasal airflow on stimulus delivery, subjects were instructed to breathe through their mouth for the duration of the task. Each trial was initiated by viewing 4 s of blank screen, followed by an on-screen 6-s countdown to the appearance of a random duration of a fixation cross (mean 4.2 s, range 2–7 s) in the center of the screen. This fixation cross remained visible during a delay of random length, making a total interstimulus interval of 12–17 s between the end of the first odor presentation and the beginning of the next odor presentation. After the fixation cross, either an odor was presented or a random word appeared in place of the fixation cross; alternatively, no odor or word appeared (blank stimulus). If an odor was presented, the fixation cross remained visible to prevent visually alerting the subject to olfactory stimulation. The odor was presented between 1 and 3 s, leading to an intertrial interval of 13–20 s. The random duration of presentation of the fixation cross was chosen in order to limit a priming effect on subjects to respond to the upcoming stimulus. Subjects were instructed to use a keyboard to indicate whether or not they detected either an odor or a word on the screen and were allowed a maximum of 3 s to respond. The visual reaction task was included to limit subjects’ focus on the odor itself. Subjects’ responses triggered termination of an olfactory or visual stimulus, and delivery of the control air flow was initiated (see Fig. 1). Subjects were presented with two blocks of 32 trials, each split evenly between 16 olfactory trials (4 per odor) and 16 visual distracter trials, resulting in 8 presentations of each odor per subject. A 5-min break was inserted between the two blocks to prevent fatigue and odor adaptation. No performance feedback was given to the subject at any time. 2.5. Statistical analysis The delay to odor delivery (450 ms) was first subtracted from all individual responses. Outliers were then identified and removed by means of two data reduction steps. First, we removed responses with reaction times shorter than 100 ms, the minimum possible reaction time. Second, we removed responses with reaction times that differed from the category mean by more than three standard deviations; this second step was not performed for the analyses of detection accuracy. Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, IL). We computed a repeated measure mixed general linear model (with a compound symmetric
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317
315
Table 1 Mean ratings, scale ranging from 0 to 10. For pleasantness, 0 indicates “very unpleasant”, 5 indicates “neutral”, and 10 indicates “very pleasant”. For intensity, 0 indicates “not perceivable” and 10 indicates “extremely strong”. For familiarity, 0 indicates “not familiar at all” and 10 indicates “very familiar”. Odor type
Odor
Pleasantness
SD
Intensity
SD
Familiarity
SD
Food
Orange Fish
7.5 3.3
1.8 2.0
3.4 4.0
1.1 2.0
7.3 5.5
2.7 3.0
Non-food
Rose Dirty socks
6.6 2.7
2.2 1.7
3.2 3.5
1.9 2.1
6.6 4.4
2.7 2.4
covariance structure), with valence (pleasant, unpleasant odor) and edibility (food, non-food odor) as within-subject variables, intensity and familiarity as covariate, and sex as a between-subject variable. Detection accuracy, the number of correct responses made within the allotted time frame, was not normally distributed, and therefore analyzed with non-parametric Wilcoxon signed rank tests, for valence and edibility. Overall evaluation of each odor was computed, combining the ratings for pleasantness, intensity, and familiarity, and correlated to reaction time (Pearson) or degree of accuracy (Spearman). Perceptual ratings of each odor were compared to a hypothetical neutral value of 5 by one-sample Student’s t-tests, and differences were assessed with a repeated measure ANOVA.
3. Results All four odors were judged to be similarly intense, F(3,117) = 1.90; p = .134, and either significantly more pleasant or unpleasant than a neutral rating, all p < .01. Furthermore, subsequent analyses demonstrated that the pleasant odors were significantly more pleasant than the unpleasant odors, F(1,39) = 165.5; p < .01. Familiarity ratings were significantly different between the four odors (F(3,117) = 9.78, p < .01) (Table 1). On average, subjects responded fastest to the fish odor (negative, food), and slowest to the rose odor (positive, non-food); mean reaction times (SD) for the individual odors were Fish 1633 ms (410), Dirty socks 1634 ms (608), Orange 1682 ms (610), Rose 1734 ms (650) (Fig. 2). The repeated measure mixed general linear model revealed a significant effect of valence on reaction times, F(1,117) = 14.4; p < .01, indicating that subjects reacted faster to the unpleasant than to the pleasant odor stimuli. Edibility also had a significant effect on reaction times, F(1,117) = 9.9; p < .01, indicating that subjects reacted faster to the food odors than to the non-food odors. Furthermore, we observed a significant interaction between both categorical factors (valence × edibility: F(1,117) = 4.7; p = .033), indicating that the accelerative effect of unpleasantness on reaction time was more evident for food odors than for non-food odors (Fig. 2). Although there was no significant difference between the odors with respect to intensity ratings, to
control for subsignificant variation herein, the above mentioned model was corrected for intensity ratings. Results remained, however, unchanged: a significant effect of valence, F(1,117) = 13.7; p < .01, and edibility, F(1,117) = 9.4; p < .01, and a significant interaction of valence × edibility, F(1,116) = 4.6; p = .035, indicating that reaction times were fastest for the negative non-food odor (fish) independent of intensity ratings. There was a significant difference between odors for familiarity ratings. To control for potential impact of familiarly ratings, as for intensity ratings, we also corrected for familiarly ratings. There was an overall significant effect of familiarity on reaction times, F(1,129) = 7.0; p < .01. However, when controlling for familiarity, the results remained unchanged: a significant effect of valence, F(1,118) = 6.0; p = .016, and edibility, F(1,116) = 13.2; p < .01, as well as a significant interaction of valence × edibility, F(1,116) = 5.4; p = .022; indicating that familiarity was not the mediating factor behind our findings. Post hoc t-tests revealed that subjects reacted much faster to the fish odor than to any other odor (all p < .01; Bonferroni corrected). The standard deviation of response times to this negative food odor was also lower than that of response times to all other odors. No significant differences between reaction times of the three other odors were observed. Attention to a stimulus is reflected not only in the speed of a reaction but also in the rate of accurate detection. An analysis of responses by odor category demonstrated that subjects accurately detected unpleasant odors before the imposed time limit more frequently than pleasant odors, z = −2.35, p = .019, and that subjects were also better at detecting food odors than non-food odors, z = −2.76, p < .01. Median detection rates (interquartile range), out of 8 trials per odor, were Fish 8 (0), Dirty socks 8 (1), Orange 8 (0), Rose 8 (1). A Friedman test revealed the highest detection accuracy for fish, a negative food odor, compared to all the other odors, 2 = 10.35, p = .016. The standard deviation of detection accuracy with the fish odor was also lower than that of detection accuracies with all other odors. In other words, subjects were more accurate in detection of the negative food odor than in detection of any other odor. No significant effect of sex on reaction time or detection accuracy was found. Reaction times for Orange and Dirty socks, but not for Fish and Rose, were significantly correlated to overall evaluation of the odors, r = .55, p < .01 for Orange; r = .37, p = .020 for Dirty socks. There were no correlations between detection rates and overall evaluation of any of the odors. 4. Discussion
Fig. 2. Mean reaction time of the individual odors measured. Grey bars indicate food odors and black bars indicate non-food odors. The * indicates that reaction times for fish odor are significantly different (p < .01) from all other odors, as demonstrated with Bonferroni corrected post hoc tests. Error bars indicate standard deviation. PEA = phenylethyl alcohol (Rose), IVA = isovaleric acid (Dirty socks).
We have demonstrated that responses to olfactory stimuli are influenced by both odor valence and edibility of the odor source. Specifically, we show that reaction time to an unpleasant food odor is faster than to other odor categories (valence, edibility). Furthermore, negative food odors are detected not only faster, but also more accurately than other odor categories, thus lending support to the hypothesis that ecological relevance influences olfactory processing. The olfactory system functions as an important screening system for both the respiratory and the gastrointestinal systems. The ingestion of spoiled food, for example, can have negative
316
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317
consequences ranging from temporarily debilitating to fatal. It is therefore important to detect these odors quickly and accurately and to react accordingly. Bensafi et al. (2002b) observed that unpleasant odors evoke greater changes in autonomic response (heart rate variation) than pleasant odors suggesting the importance of valence in adequate human responses to odor stimuli. Indeed, it is commonly argued that valence is the most important factor in odor assessment (Haddad et al., 2008; Khan et al., 2007). However, despite the aforementioned evidence for an accelerative effect of odor unpleasantness (Bensafi et al., 2002c; Jacob and Wang, 2006), contradictory data do exist. A recent study demonstrated slower – not faster – responses to a negative odor, relative to a positive odor (Chen and Dalton, 2005). The positive odor (lemon) was processed faster than a neutral odor (isopropyl alcohol), whereas the negative odor (fecal) was processed more slowly than the neutral odor. Reconciliation for these apparently conflicting results may be found in the data presented here, which demonstrate that valence and edibility both play influential roles in determining the speed of responses to olfactory stimuli. In the conflicting studies above, the pleasant odors often belonged to the food category (vanilla, fruity, lemon), and the unpleasant odors often belonged to the non-food category (valeric acid, indole, feces). Our results suggest that subjects respond faster to unpleasant odors than to pleasant odors, and faster to food odors than to non-food odors. It is important to note that the lack of correlation between reaction time and perceived pleasantness within any given odor demonstrates that the odor category, rather than the perceived valence, is the determining factor in response speed. The possibility of odor-specific effects aside, if our findings can be generalized to all odors, they suggest that the use of a food odor in the pleasant odor condition and a non-food odor in the unpleasant odor condition might lead to an overlap of valence and edibility effects. This categorical overlap would explain the contradictory findings between the pleasant and the unpleasant odor in some experimental conditions (Bensafi et al., 2002a; Chen and Dalton, 2005). Our results indicate that the ecological importance of the olfactory stimulus may play a significant role in how the brain processes the odor. The notion that stimuli conveying ecologically-relevant information are processed more accurately, as well as more quickly, than corresponding positive or neutral stimuli has been demonstrated for the visual system (Mineka and Ohman, 2002) and, here, for the olfactory system. Fish, the negative food odor, was processed with higher accuracy than the other odors. However, although we can demonstrate a clear disassociation between odor categories of edibility and pleasantness in this study, conclusions are limited by the representation of each category with a single odor. Further studies using more odorants with varying degrees of ecological relevance are warranted before finite conclusions can be made. Mineka and Ohman (2002) have described a model for fear elicitation and learning that is based on evolutionary roots. They suggest that fear-relevant stimuli have led to superior conditioning of aversive associations, when compared to fear-irrelevant stimuli, throughout evolution. These fear-relevant stimuli automatically activate specific networks in the brain (including the amygdala) before cognitive analysis of the stimulus can occur. Additional evidence for the use of separate neuronal networks for aversive stimuli comes from studies on patients, who, in spite of being cortically blind (so-called blindsight), could still be fearconditioned to visual stimuli (Hamm et al., 2003). In other words, threatening or aversive stimuli are prioritized by the cerebral system because they invariably demand faster responses. As has been shown in previous olfactory studies, odor pleasantness influences various arousal measures (Bensafi et al., 2002a; Alaoui-Ismaïli et al., 1997a,b; Brauchli et al., 1995). Combined with the abovementioned fear-model, we therefore hypothesize that the unpleasant
fish odor in the current study evokes higher arousal and attention than the other used odors, and is consequently prioritized by the brain, thus leading to faster responses and more accurate detection rates. To avoid any confounds of differential processing by the left and right hemispheres and to mimic reality as closely as possible, all odors in our study were presented birhinally. Bensafi et al. (2002a) recently demonstrated that faster responses to an unpleasant odor were dependent on the nostril of odor delivery. When the right nostril was stimulated, subjects responded faster to the unpleasant odor (indole) than to the pleasant odor (vanillin). No such difference was observed for stimulation of the left nostril. The authors interpreted these results as an indication that the right hemisphere more efficiently decodes unpleasant olfactory stimuli (Bensafi et al., 2002a). Due to birhinal odor presentation, we cannot speculate about a potential hemispheric difference. From an ecological perspective, however, it makes little sense for the olfactory system to depend on a single nostril for the detection of odors signaling a potential danger. We know that most odorants stimulate the trigeminal system in addition to the olfactory nerve (Doty et al., 1978). It is therefore possible that the observed differences between the odors are due to a separation in their degree of ability to stimulate the trigeminal system. To control for potential differences in the levels of trigeminal irritation triggered by each odorant, a birhinal localization task was conducted for each of the four odorants in a panel of eleven subjects. A birhinal localization score significantly higher than expected chance value indicates than an odorant has trigeminal properties (Hummel et al., 2003; Wysocki et al., 2003). Subjects were able to make a significant number of correct lateralization judgments, as tested with a one-sample t-test against expected chance value, for the orange odor (p < .001) but no other odor (all p > 0.1). Based on this data, we conclude that the shortened reaction time to the fish odor was not likely due to trigeminal activation. To investigate early olfactory cerebral processing with a very high temporal solution, the measurement of olfactory eventrelated potentials (ERP) is the method of choice (Pause and Krauel, 2000). Differences in ERP latency are commonly attributed to differences in the neural networks recruited for processing, and differences in ERP amplitude indicate differences in the intensity of activity within a given neural network (Kok, 1997). Although pleasant and unpleasant odors appear to evoke different electrophysiological responses (Pause and Krauel, 2000; Lundstrom et al., 2006b), there appears to be no straightforward relationship between pleasantness-dependent reaction time and ERP parameters, such as latencies. Interestingly, certain ecologically-relevant odor categories, such as endogenous odors, are processed significantly faster by the human brain, as indicated by shorter latencies (Lundstrom et al., 2006a). Whether negative food odors, another ecologically-relevant odor category, are processed more quickly by the human brain, or whether the effects reported here are reactiontime dependent, is ripe for future work. 5. Conclusion Subjects responded faster and more accurately to unpleasant odors than to pleasant odors and faster to food odors than to nonfood odors. This suggests that the olfactory system may react faster and more accurately to ecologically-relevant stimuli that warn of potential danger. References Alaoui-Ismaïli, O., Robin, O., Rada, H., Dittmar, A., Vernet-Maury, E., 1997a. Basic emotions evoked by odorants: comparison between autonomic responses and self-evaluation. Physiology & Behavior 62, 713–720.
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317 Alaoui-Ismaïli, O., Vernet-Maury, E., Dittmar, A., Delhomme, G., Chanel, J., 1997b. Odor hedonics: connection with emotional response estimated by autonomic parameters. Chemical Senses 22, 237–248. Bensafi, M., Rouby, C., Farget, V., Bertrand, B., Vigouroux, M., Holley, A., 2002a. Autonomic nervous system responses to odours: the role of pleasantness and arousal. Chemical Senses 27, 703–709. Bensafi, M., Rouby, C., Farget, V., Bertrand, B., Vigouroux, M., Holley, A., 2002b. Influence of affective and cognitive judgments on autonomic parameters during inhalation of pleasant and unpleasant odors in humans. Neuroscience Letters 319, 162–166. Bensafi, M., Rouby, C., Farget, V., Vigouroux, M., Holley, A., 2002c. Asymmetry of pleasant vs. unpleasant odor processing during affective judgment in humans. Neuroscience Letters 328, 309–313. Boesveldt, S., Hummel, T., Tessler-Lindau, S., Lundstrom, J.N., 2009. Olfactory, but not gustatory function, correlates with BMI and depressive symptoms in the elderly. Chemical Senses 34, A25. Brauchli, P., Rüegg, P.B., Etzweiler, F., Zeier, H., 1995. Electrocortical and autonomic alteration by administration of a pleasant and an unpleasant odor. Chemical Senses 20, 505–515. Chen, D., Dalton, P., 2005. The effect of emotion and personality on olfactory perception. Chemical Senses 30, 345–351. Dielenberg, R.A., McGregor, I.S., 2001. Defensive Behavior in Rats Towards Predatory Odors: A Review. Pergamon–Elsevier Science Ltd, pp. 597–609. Doty, R.L., Brugger, W.E., Jurs, P.C., Orndorff, M.A., Snyder, P.J., Lowry, L.D., 1978. Intranasal trigeminal stimulation from odorous volatiles: psychometric responses from anosmic and normal humans. Physiology & Behavior 20, 175–185. Fusari, A., Ballesteros, S., 2008. Identification of odors of edible and nonedible stimuli as affected by age and gender. Behavior Research Methods 40, 752–759. Haddad, R., Khan, R., Takahashi, Y.K., Mori, K., Harel, D., Sobel, N., 2008. A metric for odorant comparison. Nature Methods 5, 425–429. Hamm, A.O., Weike, A.I., Schupp, H.T., Treig, T., Dressel, A., Kessler, C., 2003. Affective blindsight: intact fear conditioning to a visual cue in a cortically blind patient. Brain 126, 267–275. Hansen, C.H., Hansen, R.D., 1988. Finding the face in the crowd: an anger superiority effect. Journal of Personality and Social Psychology 54, 917–924. Hummel, T., Futschik, T., Frasnelli, J., Huttenbrink, K.B., 2003. Effects of olfactory function, age, and gender on trigeminally mediated sensations: a study based on the lateralization of chemosensory stimuli. Toxicology Letters 140–141, 273–280. Jacob, T.J.C., Wang, L., 2006. A new method for measuring reaction times for odour detection at iso-intensity: comparison between an unpleasant and pleasant odour. Physiology & Behavior 87, 500–505.
317
Khan, R.M., Luk, C.H., Flinker, A., Aggarwal, A., Lapid, H., Haddad, R., Sobel, N., 2007. Predicting odor pleasantness from odorant structure: pleasantness as a reflection of the physical world. Journal of Neuroscience 27, 10015–10023. Kok, A., 1997. Event-related-potential (ERP) reflections of mental resources: a review and synthesis. Biological Psychology 45, 19–56. LeDoux, J.E., 2000. Emotion circuits in the brain. Annual Review of Neuroscience 23, 155–184. Lundstrom, J.N., Boyle, J.A., Zatorre, R.J., Jones-Gotman, M., 2008. Functional neuronal processing of body odors differ from that of similar common odors. Cerebral Cortex 18, 1466–1474. Lundstrom, J.N., Gordon, A.R., Alden, E.C., Albrecht, J., Boesveldt, S., submitted for publication. Methods for building an inexpensive computer-controlled olfactometer for temporally precise behavioral experiments. Lundstrom, J.N., Olsson, M.J., Schaal, B., Hummel, T., 2006a. A putative social chemosignal elicits faster cortical responses than perceptually similar odorants. Neuroimage 30, 1340–1346. Lundstrom, J.N., Seven, S., Olsson, M.J., Schaal, B., Hummel, T., 2006b. Olfactory event-related potentials reflect individual differences in odor valence perception. Chemical Senses 31, 705–711. Mineka, S., Ohman, A., 2002. Phobias and preparedness: the selective, automatic, and encapsulated nature of fear. Biological Psychiatry 52, 927–937. Morris, J.S., Ohman, A., Dolan, R.J., 1999. A subcortical pathway to the right amygdala mediating “unseen” fear. Proceedings of the National Academy of Sciences of the United States of America 96, 1680–1685. Ohman, A., Carlsson, K., Lundqvist, D., Ingvar, M., 2007. On the unconscious subcortical origin of human fear. Physiology & Behavior 92, 180–185. Ohman, A., Flykt, A., Esteves, F., 2001a. Emotion drives attention: detecting the snake in the grass. Journal of Experimental Psychology. General 130, 466–478. Ohman, A., Lundqvist, D., Esteves, F., 2001b. The face in the crowd revisited: a threat advantage with schematic stimuli. Journal of Personality and Social Psychology 80, 381–396. Pause, B.M., Krauel, K., 2000. Chemosensory event-related potentials (CSERP) as a key to the psychology of odors. International Journal of Psychophysiology 36, 105–122. Plailly, J., Howard, J.D., Gitelman, D.R., Gottfried, J.A., 2008. Attention to odor modulates thalamocortical connectivity in the human brain. Journal of Neuroscience 28, 5257–5267. Tooby, J., Cosmides, L., 1990. The past explains the present—emotional adaptations and the structure of ancestral environments. Ethology and Sociobiology 11, 375–424. Wysocki, C.J., Cowart, B.J., Radil, T., 2003. Nasal trigeminal chemosensitivity across the adult life span. Perception & Psychophysics 65, 115–122.
Contents lists available at ScienceDirect
Biological Psychology journal homepage: www.elsevier.com/locate/biopsycho
The fish is bad: Negative food odors elicit faster and more accurate reactions than other odors S. Boesveldt a,∗ , J. Frasnelli b , A.R. Gordon a , J.N. Lundström a,c,d a
Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA, United States CERNEC, Université de Montréal, Montréal, QC, Canada Dept. of Psychology, University of Pennsylvania, Philadelphia, PA, United States d Dept. of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden b c
a r t i c l e
i n f o
Article history: Received 4 December 2009 Accepted 7 March 2010 Available online 21 March 2010 Keywords: Reaction time Evolutionary psychology Odors Categories Valence Food
a b s t r a c t Dissociating between ‘good’ or ‘bad’ odors is arguable of crucial value for human survival, since unpleasant odors often signal danger. Therefore, negative odors demand a faster response in order to quickly avoid or move away from negative situations. We know from other sensory systems that this effect is most evident for stimuli from ecologically-relevant categories. In the olfactory system the classification of odors into the food or non-food category is of eminent importance. We therefore aimed to explore the link between odor processing speed and accuracy and odor edibility and valence by assessing response time and detection accuracy. We observed that reaction time and detection accuracy are influenced by both pleasantness and edibility. Specifically, we showed that an unpleasant food odor is detected faster and more accurately than odors of other categories. These results suggest that the olfactory system reacts faster and more accurately to ecologically-relevant stimuli that signal a potential danger. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Discrimination between ‘good’ and ‘bad’ odors is arguably of crucial value for human survival. As is well known, unpleasant odors often signal danger, such as spoiled food or a toxin, and, in some non-human species, they even serve as a warning signal of nearby predators (Dielenberg and McGregor, 2001). Negative odors demand a faster response than neutral or pleasant odors because survival depends more often on an organism’s quick response to signals of negative rather than positive situations. Two behavioral studies support this basic assumption by demonstrating that response times of human subjects to unpleasant odors were significantly shorter than for pleasant odors (Bensafi et al., 2002c; Jacob and Wang, 2006). The same principle has been demonstrated in the visual system. Hansen and Hansen showed that an angry face in a crowd of benign or happy faces was detected faster than a happy or benign face in a crowd of angry faces, suggesting that humans are more attentive to threatening signals (Hansen and Hansen, 1988). These findings were later confirmed and extended (Ohman et al., 2001b). The coupling of emotional activation and efficient capture of attention goes beyond faces, as demonstrated by findings that fear-relevant pictures of snakes and spiders were detected faster than fear-
∗ Corresponding author. Tel.: +1 267 519 4688. E-mail address: [email protected] (S. Boesveldt). 0301-0511/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.biopsycho.2010.03.006
irrelevant pictures, such as flowers and mushrooms (Ohman et al., 2001a). These and similar data have been taken as evidence that natural selection has honed the human ability to identify and react to stimuli important for survival (Mineka and Ohman, 2002; Tooby and Cosmides, 1990). Threatening stimuli, such as indicators of spoiled food or the presence of a snake, are prioritized by the cerebral system to effect faster preparation for “fight or flight” action. In the rodent brain, there appears to be a specific neural circuitry – direct linkage of the perceptual and defense systems via the thalamus and amygdala – to achieve this fast mobilization (for a review, please see LeDoux, 2000). There is evidence of a similar system in the human brain (Ohman et al., 2007), also for olfactory stimuli, since the amygdala is located only one synapse away from the olfactory receptors. Moreover, although it is widely assumed that the human olfactory system, unlike the other senses, is independent of thalamic relay, it was recently demonstrated that the thalamus indeed has a functional role in odor processing (Plailly et al., 2008). The transthalamic network is suggested to be a modulatory target of olfactory attentional processing and may serve as an attentional filter, helping to select only those inputs with behavioral relevance for processing downstream of the orbitofrontal cortex (Plailly et al., 2008). Further evidence from research on human subjects demonstrates that biologically relevant stimuli enjoy faster (Lundstrom et al., 2006a) or more direct neuronal pathways (Lundstrom et al., 2008; Morris et al., 1999), resulting in both decreased processing times and lower reaction thresholds than perceptually similar stimuli with no obvious evolutionary importance.
314
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317
Whether the human brain performs categorical olfactory processing akin to that which has been demonstrated in the visual system is not known. Collectively, food odors seem to comprise a category of particular importance to humans: our ability to identify food odors is superior to our performance with odors of nonfood objects (Boesveldt et al., 2009; Fusari and Ballesteros, 2008). Further evidence that the human olfactory system is capable of categorical discrimination comes from the previously-mentioned studies showing that reactions to negative odors are faster than to positive odors (Bensafi et al., 2002c; Jacob and Wang, 2006). Unfortunately, in both of those studies, the pleasant stimulus was a food-related odor (vanillin or amyl acetate) and the unpleasant stimulus was a non-food-related odor (indole or valeric acid), thereby making a direct comparison between odor categories – either pleasant versus unpleasant or food versus non-food – impossible. The present study sought to explore the link between odor processing speed, edibility and valence. To this end, we measured subjects’ reaction time and detection accuracy with two food odors and two non-food odors. In contrast to previous studies, each edibility category was comprised of one pleasant and one unpleasant odor. Based on what is known for the visual system, we hypothesized that, due to its ecological relevance, negative food odors would be processed faster and with higher accuracy than odors from other categories (i.e. pleasant versus unpleasant, food versus non-food). 2. Materials and methods 2.1. Participants A total of 40 young healthy participants (mean age 25 years, range 18–31 years; 20 women) were recruited via posters on the university campus. Subjects were screened prior to inclusion by means of a self-report survey for numerous nasal and neurological disorders known to affect olfactory function. None of the participating women were pregnant, and all had regular menstrual cycles of normal length. Detailed information regarding the experiment was given and written informed consent was obtained from all subjects prior to testing. All aspects of the study were performed in accordance with the University of Pennsylvania IRB and internal regulation at the Monell Chemical Senses Center. 2.2. Stimulus presentation The odors were presented with a computer-controlled 8-channel olfactometer to assure accurate odor onset and a steep odor rise-time. The premise of the olfactometer is that a valve control unit regulates the state of the olfactometer’s eight solenoid valves, each of which directs a continuous airstream of 4 liters per minute (lpm) either into the olfactometer (un-triggered state) or into an odor reservoir (triggered state). When triggered by the valve control unit, a valve directs the airstream into an odor reservoir, and the odorized headspace is transported to the birhinal nosepiece. One of the eight channels serves as a conduit for the control odor and operates in a manner identical to the other channels. Closure of an odor valve triggers the control valve such that whenever an odor is not being delivered, the control air flow is directed to the nosepiece. In the nosepiece, air carrying an odor or the control air mixes with a continuous, low-flow airstream (0.5 lpm). This continuous flow airstream masks the tactile cues that might otherwise alert the subject to channel-switching (Lundstrom et al., submitted for publication). The temporal characteristics of the odor stimulus delivered with the olfactometer settings used in this study (length of tubing, selected airflow rate, etc.) were measured with a photoionization detector (miniPID 200A, Aurora Scientific Inc., Aurora, ON, Canada). The delay to odor delivery, the lapse between the time at which a computer triggers the solenoid valve to the time at which the delivered odor reaches 90% strength, was approximately 450 ms. This calculation is based on the combined lengths of stimulus onset (400 ms, measured from computer trigger) and 10/90% odor rise-time (53 ms, from 10% to 90% odor strength). The stimulus presentation program E-Prime 2.0 Professional (Psychology Software Tools Inc., Pittsburgh, PA) was used to trigger the olfactometer, present the visual cues, and record subjects’ reaction time. 2.3. Odor stimuli Four odors were chosen on the basis of their edibility (food, non-food) and valence. The two food odorants were natural mixtures chosen to mimic the odor of the food object, and the two non-food odorants were single compounds. The food odors were orange (cold-pressed Californian orange oil, Sigma Aldrich, St. Louis,
Fig. 1. Experimental design of reaction time task and timing.
MO, USA) and fish (Fish flavor oil, Givaudan Inc., Geneva, Switzerland), which were diluted with mineral oil (70%, v/v) and 1,2-propanediol (27%, v/v), respectively. The non-food odors were rose (Phenylethyl alcohol, Sigma Aldrich, St. Louis, MO, USA) and dirty socks (Isovaleric acid, Sigma Aldrich, St. Louis, MO, USA), both of which were diluted with 1,2-propanediol to concentrations of 90% (v/v) and 40% (v/v), respectively. Pilot data indicated that these concentrations were iso-intense. Furthermore, each edibility category contained one pleasant odor (orange, rose) and one unpleasant odor (fish, dirty socks). 2.4. Experimental paradigm Perceptual ratings of pleasantness and intensity were obtained by presenting each odorant (10 ml) in individual 100 ml amber glass bottles devoid of visual markers. Odor pleasantness and familiarity were rated on a visual analogue scale (VAS, 10 cm) ranging from ‘Extremely unpleasant/unfamiliar’ to ‘Extremely pleasant/familiar’, and odor intensity was rated on a labeled magnitude scale (LMS, 10 cm) ranging from ‘No sensation’ to ‘Strongest imaginable’. Subjects were seated in a comfortable chair, fitted with the olfactometer nosepiece and in-ear headphones, oriented towards an adjustable computer monitor set at eye-level (1 m viewing distance). To exclude the possibility that auditory cues might influence subjects’ performances, low volume brown noise was played through the headphones throughout the task. To limit the influence of nasal airflow on stimulus delivery, subjects were instructed to breathe through their mouth for the duration of the task. Each trial was initiated by viewing 4 s of blank screen, followed by an on-screen 6-s countdown to the appearance of a random duration of a fixation cross (mean 4.2 s, range 2–7 s) in the center of the screen. This fixation cross remained visible during a delay of random length, making a total interstimulus interval of 12–17 s between the end of the first odor presentation and the beginning of the next odor presentation. After the fixation cross, either an odor was presented or a random word appeared in place of the fixation cross; alternatively, no odor or word appeared (blank stimulus). If an odor was presented, the fixation cross remained visible to prevent visually alerting the subject to olfactory stimulation. The odor was presented between 1 and 3 s, leading to an intertrial interval of 13–20 s. The random duration of presentation of the fixation cross was chosen in order to limit a priming effect on subjects to respond to the upcoming stimulus. Subjects were instructed to use a keyboard to indicate whether or not they detected either an odor or a word on the screen and were allowed a maximum of 3 s to respond. The visual reaction task was included to limit subjects’ focus on the odor itself. Subjects’ responses triggered termination of an olfactory or visual stimulus, and delivery of the control air flow was initiated (see Fig. 1). Subjects were presented with two blocks of 32 trials, each split evenly between 16 olfactory trials (4 per odor) and 16 visual distracter trials, resulting in 8 presentations of each odor per subject. A 5-min break was inserted between the two blocks to prevent fatigue and odor adaptation. No performance feedback was given to the subject at any time. 2.5. Statistical analysis The delay to odor delivery (450 ms) was first subtracted from all individual responses. Outliers were then identified and removed by means of two data reduction steps. First, we removed responses with reaction times shorter than 100 ms, the minimum possible reaction time. Second, we removed responses with reaction times that differed from the category mean by more than three standard deviations; this second step was not performed for the analyses of detection accuracy. Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, IL). We computed a repeated measure mixed general linear model (with a compound symmetric
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317
315
Table 1 Mean ratings, scale ranging from 0 to 10. For pleasantness, 0 indicates “very unpleasant”, 5 indicates “neutral”, and 10 indicates “very pleasant”. For intensity, 0 indicates “not perceivable” and 10 indicates “extremely strong”. For familiarity, 0 indicates “not familiar at all” and 10 indicates “very familiar”. Odor type
Odor
Pleasantness
SD
Intensity
SD
Familiarity
SD
Food
Orange Fish
7.5 3.3
1.8 2.0
3.4 4.0
1.1 2.0
7.3 5.5
2.7 3.0
Non-food
Rose Dirty socks
6.6 2.7
2.2 1.7
3.2 3.5
1.9 2.1
6.6 4.4
2.7 2.4
covariance structure), with valence (pleasant, unpleasant odor) and edibility (food, non-food odor) as within-subject variables, intensity and familiarity as covariate, and sex as a between-subject variable. Detection accuracy, the number of correct responses made within the allotted time frame, was not normally distributed, and therefore analyzed with non-parametric Wilcoxon signed rank tests, for valence and edibility. Overall evaluation of each odor was computed, combining the ratings for pleasantness, intensity, and familiarity, and correlated to reaction time (Pearson) or degree of accuracy (Spearman). Perceptual ratings of each odor were compared to a hypothetical neutral value of 5 by one-sample Student’s t-tests, and differences were assessed with a repeated measure ANOVA.
3. Results All four odors were judged to be similarly intense, F(3,117) = 1.90; p = .134, and either significantly more pleasant or unpleasant than a neutral rating, all p < .01. Furthermore, subsequent analyses demonstrated that the pleasant odors were significantly more pleasant than the unpleasant odors, F(1,39) = 165.5; p < .01. Familiarity ratings were significantly different between the four odors (F(3,117) = 9.78, p < .01) (Table 1). On average, subjects responded fastest to the fish odor (negative, food), and slowest to the rose odor (positive, non-food); mean reaction times (SD) for the individual odors were Fish 1633 ms (410), Dirty socks 1634 ms (608), Orange 1682 ms (610), Rose 1734 ms (650) (Fig. 2). The repeated measure mixed general linear model revealed a significant effect of valence on reaction times, F(1,117) = 14.4; p < .01, indicating that subjects reacted faster to the unpleasant than to the pleasant odor stimuli. Edibility also had a significant effect on reaction times, F(1,117) = 9.9; p < .01, indicating that subjects reacted faster to the food odors than to the non-food odors. Furthermore, we observed a significant interaction between both categorical factors (valence × edibility: F(1,117) = 4.7; p = .033), indicating that the accelerative effect of unpleasantness on reaction time was more evident for food odors than for non-food odors (Fig. 2). Although there was no significant difference between the odors with respect to intensity ratings, to
control for subsignificant variation herein, the above mentioned model was corrected for intensity ratings. Results remained, however, unchanged: a significant effect of valence, F(1,117) = 13.7; p < .01, and edibility, F(1,117) = 9.4; p < .01, and a significant interaction of valence × edibility, F(1,116) = 4.6; p = .035, indicating that reaction times were fastest for the negative non-food odor (fish) independent of intensity ratings. There was a significant difference between odors for familiarity ratings. To control for potential impact of familiarly ratings, as for intensity ratings, we also corrected for familiarly ratings. There was an overall significant effect of familiarity on reaction times, F(1,129) = 7.0; p < .01. However, when controlling for familiarity, the results remained unchanged: a significant effect of valence, F(1,118) = 6.0; p = .016, and edibility, F(1,116) = 13.2; p < .01, as well as a significant interaction of valence × edibility, F(1,116) = 5.4; p = .022; indicating that familiarity was not the mediating factor behind our findings. Post hoc t-tests revealed that subjects reacted much faster to the fish odor than to any other odor (all p < .01; Bonferroni corrected). The standard deviation of response times to this negative food odor was also lower than that of response times to all other odors. No significant differences between reaction times of the three other odors were observed. Attention to a stimulus is reflected not only in the speed of a reaction but also in the rate of accurate detection. An analysis of responses by odor category demonstrated that subjects accurately detected unpleasant odors before the imposed time limit more frequently than pleasant odors, z = −2.35, p = .019, and that subjects were also better at detecting food odors than non-food odors, z = −2.76, p < .01. Median detection rates (interquartile range), out of 8 trials per odor, were Fish 8 (0), Dirty socks 8 (1), Orange 8 (0), Rose 8 (1). A Friedman test revealed the highest detection accuracy for fish, a negative food odor, compared to all the other odors, 2 = 10.35, p = .016. The standard deviation of detection accuracy with the fish odor was also lower than that of detection accuracies with all other odors. In other words, subjects were more accurate in detection of the negative food odor than in detection of any other odor. No significant effect of sex on reaction time or detection accuracy was found. Reaction times for Orange and Dirty socks, but not for Fish and Rose, were significantly correlated to overall evaluation of the odors, r = .55, p < .01 for Orange; r = .37, p = .020 for Dirty socks. There were no correlations between detection rates and overall evaluation of any of the odors. 4. Discussion
Fig. 2. Mean reaction time of the individual odors measured. Grey bars indicate food odors and black bars indicate non-food odors. The * indicates that reaction times for fish odor are significantly different (p < .01) from all other odors, as demonstrated with Bonferroni corrected post hoc tests. Error bars indicate standard deviation. PEA = phenylethyl alcohol (Rose), IVA = isovaleric acid (Dirty socks).
We have demonstrated that responses to olfactory stimuli are influenced by both odor valence and edibility of the odor source. Specifically, we show that reaction time to an unpleasant food odor is faster than to other odor categories (valence, edibility). Furthermore, negative food odors are detected not only faster, but also more accurately than other odor categories, thus lending support to the hypothesis that ecological relevance influences olfactory processing. The olfactory system functions as an important screening system for both the respiratory and the gastrointestinal systems. The ingestion of spoiled food, for example, can have negative
316
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317
consequences ranging from temporarily debilitating to fatal. It is therefore important to detect these odors quickly and accurately and to react accordingly. Bensafi et al. (2002b) observed that unpleasant odors evoke greater changes in autonomic response (heart rate variation) than pleasant odors suggesting the importance of valence in adequate human responses to odor stimuli. Indeed, it is commonly argued that valence is the most important factor in odor assessment (Haddad et al., 2008; Khan et al., 2007). However, despite the aforementioned evidence for an accelerative effect of odor unpleasantness (Bensafi et al., 2002c; Jacob and Wang, 2006), contradictory data do exist. A recent study demonstrated slower – not faster – responses to a negative odor, relative to a positive odor (Chen and Dalton, 2005). The positive odor (lemon) was processed faster than a neutral odor (isopropyl alcohol), whereas the negative odor (fecal) was processed more slowly than the neutral odor. Reconciliation for these apparently conflicting results may be found in the data presented here, which demonstrate that valence and edibility both play influential roles in determining the speed of responses to olfactory stimuli. In the conflicting studies above, the pleasant odors often belonged to the food category (vanilla, fruity, lemon), and the unpleasant odors often belonged to the non-food category (valeric acid, indole, feces). Our results suggest that subjects respond faster to unpleasant odors than to pleasant odors, and faster to food odors than to non-food odors. It is important to note that the lack of correlation between reaction time and perceived pleasantness within any given odor demonstrates that the odor category, rather than the perceived valence, is the determining factor in response speed. The possibility of odor-specific effects aside, if our findings can be generalized to all odors, they suggest that the use of a food odor in the pleasant odor condition and a non-food odor in the unpleasant odor condition might lead to an overlap of valence and edibility effects. This categorical overlap would explain the contradictory findings between the pleasant and the unpleasant odor in some experimental conditions (Bensafi et al., 2002a; Chen and Dalton, 2005). Our results indicate that the ecological importance of the olfactory stimulus may play a significant role in how the brain processes the odor. The notion that stimuli conveying ecologically-relevant information are processed more accurately, as well as more quickly, than corresponding positive or neutral stimuli has been demonstrated for the visual system (Mineka and Ohman, 2002) and, here, for the olfactory system. Fish, the negative food odor, was processed with higher accuracy than the other odors. However, although we can demonstrate a clear disassociation between odor categories of edibility and pleasantness in this study, conclusions are limited by the representation of each category with a single odor. Further studies using more odorants with varying degrees of ecological relevance are warranted before finite conclusions can be made. Mineka and Ohman (2002) have described a model for fear elicitation and learning that is based on evolutionary roots. They suggest that fear-relevant stimuli have led to superior conditioning of aversive associations, when compared to fear-irrelevant stimuli, throughout evolution. These fear-relevant stimuli automatically activate specific networks in the brain (including the amygdala) before cognitive analysis of the stimulus can occur. Additional evidence for the use of separate neuronal networks for aversive stimuli comes from studies on patients, who, in spite of being cortically blind (so-called blindsight), could still be fearconditioned to visual stimuli (Hamm et al., 2003). In other words, threatening or aversive stimuli are prioritized by the cerebral system because they invariably demand faster responses. As has been shown in previous olfactory studies, odor pleasantness influences various arousal measures (Bensafi et al., 2002a; Alaoui-Ismaïli et al., 1997a,b; Brauchli et al., 1995). Combined with the abovementioned fear-model, we therefore hypothesize that the unpleasant
fish odor in the current study evokes higher arousal and attention than the other used odors, and is consequently prioritized by the brain, thus leading to faster responses and more accurate detection rates. To avoid any confounds of differential processing by the left and right hemispheres and to mimic reality as closely as possible, all odors in our study were presented birhinally. Bensafi et al. (2002a) recently demonstrated that faster responses to an unpleasant odor were dependent on the nostril of odor delivery. When the right nostril was stimulated, subjects responded faster to the unpleasant odor (indole) than to the pleasant odor (vanillin). No such difference was observed for stimulation of the left nostril. The authors interpreted these results as an indication that the right hemisphere more efficiently decodes unpleasant olfactory stimuli (Bensafi et al., 2002a). Due to birhinal odor presentation, we cannot speculate about a potential hemispheric difference. From an ecological perspective, however, it makes little sense for the olfactory system to depend on a single nostril for the detection of odors signaling a potential danger. We know that most odorants stimulate the trigeminal system in addition to the olfactory nerve (Doty et al., 1978). It is therefore possible that the observed differences between the odors are due to a separation in their degree of ability to stimulate the trigeminal system. To control for potential differences in the levels of trigeminal irritation triggered by each odorant, a birhinal localization task was conducted for each of the four odorants in a panel of eleven subjects. A birhinal localization score significantly higher than expected chance value indicates than an odorant has trigeminal properties (Hummel et al., 2003; Wysocki et al., 2003). Subjects were able to make a significant number of correct lateralization judgments, as tested with a one-sample t-test against expected chance value, for the orange odor (p < .001) but no other odor (all p > 0.1). Based on this data, we conclude that the shortened reaction time to the fish odor was not likely due to trigeminal activation. To investigate early olfactory cerebral processing with a very high temporal solution, the measurement of olfactory eventrelated potentials (ERP) is the method of choice (Pause and Krauel, 2000). Differences in ERP latency are commonly attributed to differences in the neural networks recruited for processing, and differences in ERP amplitude indicate differences in the intensity of activity within a given neural network (Kok, 1997). Although pleasant and unpleasant odors appear to evoke different electrophysiological responses (Pause and Krauel, 2000; Lundstrom et al., 2006b), there appears to be no straightforward relationship between pleasantness-dependent reaction time and ERP parameters, such as latencies. Interestingly, certain ecologically-relevant odor categories, such as endogenous odors, are processed significantly faster by the human brain, as indicated by shorter latencies (Lundstrom et al., 2006a). Whether negative food odors, another ecologically-relevant odor category, are processed more quickly by the human brain, or whether the effects reported here are reactiontime dependent, is ripe for future work. 5. Conclusion Subjects responded faster and more accurately to unpleasant odors than to pleasant odors and faster to food odors than to nonfood odors. This suggests that the olfactory system may react faster and more accurately to ecologically-relevant stimuli that warn of potential danger. References Alaoui-Ismaïli, O., Robin, O., Rada, H., Dittmar, A., Vernet-Maury, E., 1997a. Basic emotions evoked by odorants: comparison between autonomic responses and self-evaluation. Physiology & Behavior 62, 713–720.
S. Boesveldt et al. / Biological Psychology 84 (2010) 313–317 Alaoui-Ismaïli, O., Vernet-Maury, E., Dittmar, A., Delhomme, G., Chanel, J., 1997b. Odor hedonics: connection with emotional response estimated by autonomic parameters. Chemical Senses 22, 237–248. Bensafi, M., Rouby, C., Farget, V., Bertrand, B., Vigouroux, M., Holley, A., 2002a. Autonomic nervous system responses to odours: the role of pleasantness and arousal. Chemical Senses 27, 703–709. Bensafi, M., Rouby, C., Farget, V., Bertrand, B., Vigouroux, M., Holley, A., 2002b. Influence of affective and cognitive judgments on autonomic parameters during inhalation of pleasant and unpleasant odors in humans. Neuroscience Letters 319, 162–166. Bensafi, M., Rouby, C., Farget, V., Vigouroux, M., Holley, A., 2002c. Asymmetry of pleasant vs. unpleasant odor processing during affective judgment in humans. Neuroscience Letters 328, 309–313. Boesveldt, S., Hummel, T., Tessler-Lindau, S., Lundstrom, J.N., 2009. Olfactory, but not gustatory function, correlates with BMI and depressive symptoms in the elderly. Chemical Senses 34, A25. Brauchli, P., Rüegg, P.B., Etzweiler, F., Zeier, H., 1995. Electrocortical and autonomic alteration by administration of a pleasant and an unpleasant odor. Chemical Senses 20, 505–515. Chen, D., Dalton, P., 2005. The effect of emotion and personality on olfactory perception. Chemical Senses 30, 345–351. Dielenberg, R.A., McGregor, I.S., 2001. Defensive Behavior in Rats Towards Predatory Odors: A Review. Pergamon–Elsevier Science Ltd, pp. 597–609. Doty, R.L., Brugger, W.E., Jurs, P.C., Orndorff, M.A., Snyder, P.J., Lowry, L.D., 1978. Intranasal trigeminal stimulation from odorous volatiles: psychometric responses from anosmic and normal humans. Physiology & Behavior 20, 175–185. Fusari, A., Ballesteros, S., 2008. Identification of odors of edible and nonedible stimuli as affected by age and gender. Behavior Research Methods 40, 752–759. Haddad, R., Khan, R., Takahashi, Y.K., Mori, K., Harel, D., Sobel, N., 2008. A metric for odorant comparison. Nature Methods 5, 425–429. Hamm, A.O., Weike, A.I., Schupp, H.T., Treig, T., Dressel, A., Kessler, C., 2003. Affective blindsight: intact fear conditioning to a visual cue in a cortically blind patient. Brain 126, 267–275. Hansen, C.H., Hansen, R.D., 1988. Finding the face in the crowd: an anger superiority effect. Journal of Personality and Social Psychology 54, 917–924. Hummel, T., Futschik, T., Frasnelli, J., Huttenbrink, K.B., 2003. Effects of olfactory function, age, and gender on trigeminally mediated sensations: a study based on the lateralization of chemosensory stimuli. Toxicology Letters 140–141, 273–280. Jacob, T.J.C., Wang, L., 2006. A new method for measuring reaction times for odour detection at iso-intensity: comparison between an unpleasant and pleasant odour. Physiology & Behavior 87, 500–505.
317
Khan, R.M., Luk, C.H., Flinker, A., Aggarwal, A., Lapid, H., Haddad, R., Sobel, N., 2007. Predicting odor pleasantness from odorant structure: pleasantness as a reflection of the physical world. Journal of Neuroscience 27, 10015–10023. Kok, A., 1997. Event-related-potential (ERP) reflections of mental resources: a review and synthesis. Biological Psychology 45, 19–56. LeDoux, J.E., 2000. Emotion circuits in the brain. Annual Review of Neuroscience 23, 155–184. Lundstrom, J.N., Boyle, J.A., Zatorre, R.J., Jones-Gotman, M., 2008. Functional neuronal processing of body odors differ from that of similar common odors. Cerebral Cortex 18, 1466–1474. Lundstrom, J.N., Gordon, A.R., Alden, E.C., Albrecht, J., Boesveldt, S., submitted for publication. Methods for building an inexpensive computer-controlled olfactometer for temporally precise behavioral experiments. Lundstrom, J.N., Olsson, M.J., Schaal, B., Hummel, T., 2006a. A putative social chemosignal elicits faster cortical responses than perceptually similar odorants. Neuroimage 30, 1340–1346. Lundstrom, J.N., Seven, S., Olsson, M.J., Schaal, B., Hummel, T., 2006b. Olfactory event-related potentials reflect individual differences in odor valence perception. Chemical Senses 31, 705–711. Mineka, S., Ohman, A., 2002. Phobias and preparedness: the selective, automatic, and encapsulated nature of fear. Biological Psychiatry 52, 927–937. Morris, J.S., Ohman, A., Dolan, R.J., 1999. A subcortical pathway to the right amygdala mediating “unseen” fear. Proceedings of the National Academy of Sciences of the United States of America 96, 1680–1685. Ohman, A., Carlsson, K., Lundqvist, D., Ingvar, M., 2007. On the unconscious subcortical origin of human fear. Physiology & Behavior 92, 180–185. Ohman, A., Flykt, A., Esteves, F., 2001a. Emotion drives attention: detecting the snake in the grass. Journal of Experimental Psychology. General 130, 466–478. Ohman, A., Lundqvist, D., Esteves, F., 2001b. The face in the crowd revisited: a threat advantage with schematic stimuli. Journal of Personality and Social Psychology 80, 381–396. Pause, B.M., Krauel, K., 2000. Chemosensory event-related potentials (CSERP) as a key to the psychology of odors. International Journal of Psychophysiology 36, 105–122. Plailly, J., Howard, J.D., Gitelman, D.R., Gottfried, J.A., 2008. Attention to odor modulates thalamocortical connectivity in the human brain. Journal of Neuroscience 28, 5257–5267. Tooby, J., Cosmides, L., 1990. The past explains the present—emotional adaptations and the structure of ancestral environments. Ethology and Sociobiology 11, 375–424. Wysocki, C.J., Cowart, B.J., Radil, T., 2003. Nasal trigeminal chemosensitivity across the adult life span. Perception & Psychophysics 65, 115–122.