Physiology & Behavior, Vol. 66, No. 4, pp. 689693, 1999 1999 Elsevier Science Inc.

. Printed in the USA. All rights reserved 0031-9384/99/$see front matter

PII S0031-9384(99)00010-4

A Demonstration of Classical Conditioning of the Human Eyeblink to an Olfactory Stimulus

ANNA BACON MOORE* AND CLAIRE MURPHY1 *San Diego State University/University of CaliforniaSan Diego, Joint Doctoral Program in Clinical Psychology, and University of California Medical CenterSan Diego, 6363 Alvarado Court, Ste. 101, Department of Psychology, San Diego State University, San Diego, CA 92120-4913 Received 12 June 1998; Accepted 14 December 1998
MOORE, A. B. AND C. MURPHY. A demonstration of classical conditioning of the human eyeblink to an olfactory stimulus. PHYSIOL BEHAV 66(4) 689693, 1999.While acquisition of the eyeblink conditioned response to a variety of stimuli has been widely studied, it has yet to be established that humans will demonstrate a conditioned response to an olfactory stimulus. In this study we present data to show that humans will demonstrate a classically conditioned eyeblink response to an olfactory stimulus. Ten participants were tested in a delay conditioning procedure with an olfactory stimulus presented in a heated, humidified stream of air via an olfactometer, allowing the precise control over stimulus duration necessary for delay conditioning. Trials on which odor alone was presented were administered to four additional participants. Establishing that humans will demonstrate an eyeblink conditioned response to an olfactory stimulus will allow further exploration of the pathways involved in classical conditioning and associative learning, as well as an analysis of conditioning pathways across sensory modalities. 1999 Elsevier Science Inc. Classical conditioning Olfaction Learning Memory

CLASSICAL conditioning represents one of the most basic forms of learning and memory in which a neutral stimulus is paired with a stimulus known to produce a specific behavioral response, typically a defensive or protective response. The most widely researched and understood example of classical conditioning is that of the eyeblink conditioned response (EBCC). Classical conditioning of an eyeblink has been demonstrated in two paradigms: delay and trace (20,21,29,31). In a delay paradigm, the conditioned stimulus is presented for a set amount of time alone, then together with the unconditioned stimulus, and finally the two coterminate (See Fig. 1a). In a trace paradigm, the conditioned stimulus is presented alone for a period of time, there is then a gap, or trace period, during which no stimuli are presented; and finally, the unconditioned stimulus is presented for some time, alone (See Fig. 1b). Advantages to classical conditioning include 1) low task demand minimum effort required on the part of the participant; 2) ability to rule out sensorimotor deficits, fatigue, and motivational differences; and 3) the behavioral response is unambiguous and clearly defined (20,29). Finally, the tremendous similarities between animal, specifically rabbit, and human performance on conditioning tasks have allowed signifi1To

cant advances in understanding not only the impact of test parameters and psychometric manipulations on rates of conditioning, but also the neural substrates and biological bases of conditioning (21,31,32). Although there is a great deal of literature on performance of animals and humans in eyeblink classical conditioning to auditory stimuli (21,2426,34,35), there are, to date, no studies of eyeblink classical conditioning to olfactory stimuli. Establishing that humans can be conditioned to an olfactory stimulus in an eyeblink classical conditioning paradigm will allow further exploration of the pathways involved in classical conditioning and associative learning, as well as an analysis of conditioning pathways across sensory modalities. In addition, such a finding could prove useful in further elucidating the olfactory pathway and higher level processing of olfactory information. There is evidence, both in animal and human studies, documenting learned associations to odorant stimuli, suggesting that acquisition of a conditioned response to an olfactory stimulus is potentially possible (7,8,10,11,13,17). Given studies such as these, examination of eyeblink classical conditioning to olfactory stimuli presents as an important component in the growing literature on associative learning to odorant stimuli.

whom requests for reprints should be addressed. E-mail: cmurphy@sunstroke.sdsu.edu

689

690

MOORE AND MURPHY tioned stimulus (4). Although the amnesic participants in this study demonstrated intact delay conditioning, other studies including participants with hippocampal damage show impaired delay conditioning (20,22,3537). In addition, both imaging and electrophysiological studies show hippocampal activation during delay conditioning tasks (1,2,5). Therefore, it is hypothesized that the hippocampus plays a modulatory role in delay conditioning, but may not be essential for acquisition of the conditioned response (23,37).
METHOD

Participants Ten adults, 4 men and 6 women, ranging in age from 21 to 60 participated in this study; the mean age of participants was 32.3 years ( 15.2). The participants had an average of 13.9 years of education ( 1.5). None of the participants had a history of substance dependence, psychiatric, or neurologic illness, nor did any participants report a history of head trauma resulting in loss of consciousness. Although history of past tobacco use was not recorded, none of the participants endorsed current use of tobacco products. Informed consent was obtained from all participants prior to beginning the study. Stimuli The odorant geraniol was used in this study as the conditioned stimulus, presented in pure form, delivered via an olfactometer (described below) at a concentration of 15 ppm. This odorant was selected because it does not stimulate the trigeminal nerve (6), nor does it contaminate or adhere to surfaces in the olfactometer. To ensure that participants remained awake and alert with the eyes open, the Walt Disney movie The Lion King was shown during the study. Study Apparatus The olfactory stimulus was presented to the participant via an olfactometer (18). The participant was seated in an examination chair with a nasal canula (3-mm inner diameter) resting just inside the nostril. A steady stream of air flows through the olfactometer and out the nose piece into the participants nostril. An air compressor located outside the testing room supplies breathing air to the olfactometer. Prior to reaching the nasal canula and the participant, the air stream is heated to body temperature and humidified. This heating and humidification of the air is accomplished via air flow valves that direct the air into a washing bottle filled with warm deionized water and heated by electrically resistive heating tape until the air reaches at least 80% relative humidity and a temperature of 36.5 C. An additional washing bottle, filled with geraniol in its pure liquid form, supplies the odor vapor for the olfactometer. Release of odor vapor into the steady stream of heated and humidified air is accomplished by solenoid valves. When triggered, the solenoid valves open for a specified amount of time, allowing a portion of odor vapor flow to replace the main air flow and exit via the nasal piece resting in the subjects nostril. In addition, these solenoid valves allow the odorant to be exhausted away from the olfactometer to a vent located outside the testing room. In this manner, precise control over odor presentation is maintained and air flow rate into the nose remains constant throughout the entire length of the study. Finally, the solenoid valves are acoustically isolated to prohibit any auditory cues.

FIG. 1. Schematic of two classical conditioning paradigms. (a) Delay conditioning paradigm. (b) Trace conditioning paradigm.

It is our belief that use of an olfactory conditioned stimulus in the eyeblink classical conditioning paradigm has yet to be explored due to technological limitations. Specifically, precise control over conditioned stimulus onset and offset is critical; this is inherently difficult when working with odorant stimuli. It is our goal to demonstrate that through the use of an olfactometer designed to deliver discrete amounts of an odorant stimulus with an onset and offset that can be controlled within milliseconds, humans will demonstrate a conditioned response to an olfactory stimulus. The cerebellum is the primary site of formation and storage of the conditioned response; of particular importance are the dentate and interpositus nuclei (3,19,30,33,34). Ablation of these deep nuclei or of the middle cerebellar peduncle, the primary pathway for cerebellar afferents, abolishes the conditioned response, regardless of whether the CS modality is tone or light (15,27,28). Perhaps the most striking evidence for the importance of the cerebellum in conditioning comes from animal studies in which conditioned rabbits are decerebrated, and yet maintain the conditioned response (16); removal of the cerebellum abolishes the conditioned response in conditioned rabbits (30). The hippocampus is the other primary brain structure believed to play a role in classical conditioning; however, the role of the hippocampus is not as direct as that of the cerebellum. There is evidence to suggest that the hippocampus is essential for both acquisition and retention of the conditioned response in a trace paradigm (14,35). In a study comparing performance of amnesics with documented hippocampal damage and healthy controls on both a delay and a trace conditioning task (4), amnesic patients demonstrate intact delay conditioning, but fail to acquire trace conditioning. Not only does this finding support the theory that the hippocampus is critically involved in trace conditioning, but also strengthens and extends the theory by providing rationale and explanation for the role of the hippocampus in trace conditioning by addressing the issue of participant awareness of the relationship between the conditioned stimulus and unconditioned stimulus. Clark and Squire report that those participants in the trace conditioning paradigm who did not gain conscious knowledge of the relationship between the conditioned stimulus and unconditioned stimulus failed to condition, leading to the conclusion that the role of the hippocampus in trace conditioning is to establish the memory trace, or representation, of the relationship between the conditioned stimulus and uncondi-

OLFACTORY CLASSICAL CONDITIONING

691 tudes for all five eyeblinks in response to air alone determined an individuals eyeblink amplitude). The following instructions were read to all participants:
Our objective in this study is to examine how well peoples memory works when they are in situations that could be distracting. Therefore, I am going to be showing you part of a movie. I want you to pay careful attention to the movie because I will be asking you several questions about it when we are finished. However, I am going to try to distract you from the movie. First, I would like you to place this plastic tip just inside your nose. Air will come out of here and from time to time you will smell some odors. Next, I am going to have you wear these special glasses. Please relax and make your self comfortable. Occasionally you will feel some mild puffs of air in your eye. If you feel like blinking, please do so. Just let your natural reactions take over. Remember, try to focus on the movie and remember as many details as you can.

FIG. 2. Mean number of conditioned responses across blocks of paired CS US presentations (10 trials of CS US pairings per block).

The eyeblink apparatus consists of a low-power infrared photobeam set and a small-diameter air-puff tube attached to the left lens of a lightweight pair of sunglasses (San Diego Instruments). The photobeam measures the amount of light reflected as the eyelid closes, and this measurement is recorded as a voltage change that is computer analyzed. The air puff tube delivers 3 psi of air for 100 ms. A customized computer program (LabVIEW) was used to control the timing of the olfactory stimulus, the air puff to the eye, and recording of the photo beam reflected from the eye. Procedure The glasses holding the airpuff jet and infrared eyeblink monitor were placed on the participant. Five conditioned stimulus (odorant) presentations were administered to serve as a baseline and allow verification that the participants would not demonstrate a blink response to the odorant alone. Five unconditioned stimulus (airpuff) presentations were administered so that amplitude for a complete blink could be determined for each participant (simple averaging of the ampli-

Because the delay conditioning procedure is associated with faster acquisition of the conditioned response and a stronger response than the trace procedure, this study employed a delay conditioning paradigm [see Fig. 1a; (14)]. The delay interval between the conditioned stimulus and unconditioned stimulus (interstimulus interval; ISI) was 750 ms. Therefore, the odor was presented for 750 ms alone, then presented for an additional 100 ms in conjunction with the airpuff, and finally, the odor and airpuff coterminated. This sequence of events, lasting a total of 850 ms, constituted one trial. Ten trials constituted one block of trials. A total of 90 trials, or 9 blocks, of conditioned stimulus unconditioned stimulus pairings were presented to a subject during the course of a testing session. To avoid the problem of adaptation in the olfactory system, the intertrial interval (ITI) was 40 s, and the interval between blocks of trials was 90 to 120 s (9).
RESULTS

A conditioned response (CR) was defined as any blink occurring 150 ms after onset of the conditioned stimulus (odor) and prior to onset of the unconditioned stimulus (airpuff). Responses generated within the first 150 ms after CS onset (alpha responses) were not considered as conditioned responses because a participant could not have perceived, recognized, and responded to the conditioned stimulus within

FIG. 3. Mean number of conditioned responses across blocks of paired CS US presentations (solid line) vs. mean number of blink responses across presentations of odor alone (dashed line).

692 150 ms (12). In addition, a blink amplitude exceeded 10% of the amplitude of the average blink obtained during the first five US-alone presentations in order to meet criteria for a conditioned response. To rule out pseudoconditioning, eyeblink recordings during the initial five odor-only trials were analyzed to determine whether or not the participants demonstrated a blink response to the odor alone. The participants generated a mean of 0.9 blink responses during the five odor-only presentations (for a mean of 4.1 nonblink responses). A paired t-test confirms a significant difference between the number of blink versus nonblink responses when the odorant stimulus was presented without the air puff (p 0.01). As can be seen in Fig. 2, the data show that the mean number of conditioned responses generated during the first block of conditioned stimulus unconditioned stimulus pairings was 1.6, and that by the final block of conditioned stimulus unconditioned stimulus presentations, the participants were generating an average of 4.4 CRs, t(10) 4.58, p 0.001. During the final block of paired conditioned stimulus unconditioned stimulus presentations, participants exhibited a learned blink response to the odorant 44% of the time the odorant was presented. This increase in the number of conditioned responses generated across trials reflects the learned association between the odorant stimulus and the airpuff. Despite the fact that the geraniol concentration in this study (15 ppm) was below the threshold for trigeminal responses (6), it could be argued that repetitive administration of the odorant stimulus led to an increase in the frequency of eyeblinks. To examine this possibility, an additional four participants were tested in an odor-only paradigm. This group consisted of two men and two women, ranging in age from 25 to 35; the mean age of these participants was 27.8 ( 4.9), with an average of 14.0 ( 2.3) years of education. Inclusion criteria were identical to those described previously. The study methods were exactly as those described above, with the modification that there were no paired odor airpuff presentations. A total of nine blocks, or 90 trials, of the odor alone were presented to a subject during the course of a testing session. The intertrial interval was 40 s, and the interval between block of trials was 90 to 120 s. Following administration of the 90 odoronly trials, five air-only trials were administered to determine the average blink amplitude for each participant. During the odor-only presentations, any blink amplitude exceeding 10% of the amplitude of the average blink obtained during the five airpuff-only (unconditioned stimulus) was considered a blink response to the odorant stimulus. As can be seen in Fig. 3, the data show that the mean number of blink responses to presentation of the odorant do not significantly increase across presentations of the odor, arguing against the possibility that repetitive administration of the odor led to an increase in the frequency of eyeblinks. A t-test confirms that there is not a significant difference in the number of blink responses generated in block 1 versus block 9 (p 0.79).
DISCUSSION

MOORE AND MURPHY memory, and to examine neural associations and pathways linking not only basic reflexive behaviors, but higher level cognitive functions as well. The results of this study demonstrate that people will acquire a conditioned response to an olfactory stimulus in a delay conditioning paradigm. Participants in this study were recruited for a study on memory and peoples ability to concentrate in situations that could be distracting; the instructions read to participants greatly emphasized the importance of concentrating on the movie and not allowing extraneous stimuli to be distracting. It is possible that participants in this study would have shown even greater and more rapid conditioned response acquisition if the instructions and study recruitment had not emphasized the memory test as extensively. Studies from auditory eyeblink classical conditioning have demonstrated that manipulations on the interstimulus interval (ISI) impact the rate of conditioned response acquisition in a curvilinear manner, with younger adults showing optimal performance in a shorter ISI (400 ms). This study tested all participants, young and old, in a long ISI (750 ms). Not only is the ISI of interest, but the intertrial interval (ITI), or time between paired CS US presentations, may also prove to be an important variable. This issue is particularly complex for studies of olfaction due to neuronal adaptation to the odorant stimulus (9). A related issue concerns the fact that in auditory eyeblink classical conditioning, the conditioned stimulus intensity affects conditioning, such that the louder the tone, the greater the rate of conditioned response acquisition. Studies to further explore what impact odorant concentration has on eyeblink classical conditioning are needed. Finally, while this study establishes that humans can acquire an eyeblink conditioned response to an odorant, it will be interesting to examine rates of conditioning to a variety of odorants. Geraniol was chosen for this study because it has no trigeminal impact. Variations in the trigeminal olfactory impact of an odorant stimulus may well affect the rate and strength of the conditioned response because the neural substrates of olfactory and trigeminal processing show sensory interaction of central origin. Establishing that an odorant is an effective conditioned stimulus provides opportunity for studies designed to delineate pathways between the areas of the brain that allow us to learn and form associations and the areas of the brain that process olfactory information. This pathway has been largely explained for auditory stimuli (30). Although we know that humans form strong associations with odors, we do not, as of yet, know how these associations are formed or where they are stored. Furthermore, given the hedonic facet of olfactory stimuli, future studies of classical conditioning to odors may be able to address what role hedonics play in associative learning and memory.
ACKNOWLEDGEMENTS

Despite the fact that classical conditioning represents one of the simplest models of learning, this paradigm affords extensive opportunity to study neural substrates of learning and

This work was supported by the National Institute of Health Grants #AGO8203, #AG04085, and #DC02064 to CM. The authors wish to acknowledge and thank Drs. Paul Solomon and Robert Clark for their encouragement and expert advice throughout the design and implementation of this study, Ken Muffai for his outstanding assistance with computer programming, and Charlie Morgan for his superlative technical and olfactometric expertise.

REFERENCES 1. Blaxton, T. A.; Zeffiro, T. A.; Gabrieli, J. D.; Bookheimer, S. Y.; Carrillo, M. C.; Theodore, W. H.; Disterhoft, J. F.: Functional mapping of human learning: A positron emission tomography activation study of eyeblink conditioning. J. Neurosci. 16:40324040; 1996.

OLFACTORY CLASSICAL CONDITIONING

2. Berger, T. W.; Rinaldi, P. C.; Weisz, D. J.; Thompson, R. F.: Singleunit analysis of different hippocampal cell types during classical conditioning of rabbit nictating membrane response. J. Neurophysiol. 50:11971219; 1983. 3. Clark, R. E.; Lavond, D. G.: Neural activity in the trigeminal complex with interpositus or red nucleus inactivation during classical eyeblink conditioning. Behav. Neurosci. 110:1321; 1996. 4. Clark, R. E.; Squire, L. R.: Classical conditioning and brain systems: The role of awareness. Science 280:7781; 1998. 5. Disterhoft, J. F.; Coulter, D. A.; Alkon, D. L.: Conditioning-specific membrane changes of rabbit hippocampal neurons measured in vitro. Proc. Natl. Acad. Sci. USA 83:27332737; 1986. 6. Doty, R. L.; Brugger, W. E.; Jurs, P. C.; Orndorff, M. A.; Snyder, P. J.; Lowry, L. D.: Intranasal trigeminal stimulation from odorous volatiles: Psychometric responses from anosmic and normal humans. Physiol. Behav. 20:175185; 1978. 7. Ehrlichman, H.; Kuhl, S. B.; Zhu, J.; Warrenburg, S.: Startle reflex modulation by pleasant and unpleasant odors in a between-subjects design. Pyschophysiology 34:726729; 1997. 8. Ehrlichman, H.; Kuhl, S. B.; Zhu, J.; Warrenburg, S.: Startle reflex modulation during exposure to pleasant and unpleasant odors. Pyschophysiology 32:150154; 1995. 9. Ekman, G.; Berglund, B.; Berflund, U.: Perceived intensity of odor as a function of time of adaptation. Scand. J. Physiol. 8:177 186; 1967. 10. Fan, R. J.; Hanson, B. S.: Olfactory memory and conditional discrimination in the moth Spodoptera littoralis (Lepidoptera: Noctuidae). Soc. Neurosci. Abstr. 22: 1997. 11. Freeman, W. J.; Grajski, K. A.: Relation of olfactory EEG to behavior: Factor analysis. Behav. Neurosci. 101:766777; 1987. 12. Gormezano, I.: Classical conditioning. In: Sidowski, J. B., ed. Experimental methods and instrumentation in psychology. New York: McGraw-Hill; 1966:385420. 13. Herz, R. S.; Cupchik, G. C.: The emotional distinctiveness of odor-evoked memories. Chem. Senses 20:517528; 1995. 14. Kim, J. J.; Clark, R. E.; Thompson, R. F.: Hippocampectomy impairs the memory of recently, but not remotely, acquired trace eyeblink conditioned responses. Behav. Neurosci. 109:195203; 1995. 15. Lewis, J. L.; LoTurco, J. J.; Solomon, P. R.: Lesions of the middle cerebellar peduncle disrupt acquisition and retention of the rabbits classically conditioned nictitating membrane response. Behav. Neurosci. 101:151157; 1987. 16. Mauk, M. D.; Thompson, R. F.: Retention of classically conditioned eyelid responses following acute decerebration. Brain Res. 403:8995; 1987. 17. Meyers, C. E.; Gluck, M. A.: Cortico-hippocampal representations in simultaneous odor discrimination: A computational interpretation of Eichenbaum, Matthews, and Cohen (1989). Behavior. Neurosci. 110:685706; 1996. 18. Murphy, C.; Nordin, S.; de Wijk, R. A.; Cain, W. S.; Polich, J.: Olfactory-evoked potentials: Assessment of young and elderly, and comparison to psychophysical threshold. Chem. Senses 19: 4756; 1994. 19. Nordholm, A. F.; Thompson, J. K.; Dersarkissian, C.; Thompson, R. F.: Lidocaine infusion in a critical region of cerebellum completely prevents learning of the conditioned eyeblink response. Behav. Neurosci. 107:882886; 1993. 20. Solomon, P. R.; Levine, E.; Bein, T.; Pendlebury, W. W.: Disrup-

693
tion of classical conditioning in patients with Alzheimers disease. Neurobiol. Aging 12:283287; 1991. Solomon, P. R.; Beal, M. F.; Pendlebury, W. W.: Age-related disruption of classical conditioning: A model systems approach to memory disorders. Neurobiol. Aging 9:535546; 1988. Solomon, P. R.; Solomon, S. D.; Schaaf, E. V.; Perry, E. H.: Altered activity in the hippocampus is more detrimental to classical conditioning than removing the structure. Science 220:329 330; 1983. Solomon, P. R.; Brett, M.; GrocciaEllison, M. E.; Oyler, C.; Tomasi, M.; Pendlebury, W. W.: Classical conditioning in patients with Alzheimers disease: A multiday study. Psychol. Aging 10:248 254; 1995. Solomon, P. R.; GrocciaEllison, M.; Levine, E.; Blanchard, S.; Pendlebury, W. W.: Do temporal relationships in conditioning change across the life span? Perspectives from eyeblink conditioning in humans and rabbits. Ann. NY Acad. Sci. 608:212238; 1990. Solomon, P. R.; Pendlebury, W. W.: A model systems approach to the study of disorders of learning and memory. Neurobiol. Aging 15:283286; 1994. Solomon, P. R.; Pomerleau, D.; Bennett, L.; James, J.; Morse, D. L.: Acquisition of the classically conditioned eyeblink response in humans over the life span. Psychol. Aging 4:3441; 1989. Steinmetz, J. E.; Lavond, D. G.; Ivkovich, D.; Logan, C. G.; Thompson, R. F.: Disruption of classical eyelid conditioning after cerebellar lesions: Damage to a memory trace system or a simple performance deficit. J. Neurosci. 12:44034426; 1992 Steinmetz, J. E.; Sears, L. L.; Gabriel, M.; Kubota, Y.; Poremba, A.; Kang, E.: Cerebellar iterpositus nucleus lesions disrupt classical nictitating membrane conditioning but not discriminative avoidance learning in rabbits. Behav. Brain Res. 45:7180; 1991. Thompson, R. F.: Classical conditioning: The Rosetta Stone for brain substrates of age-related deficits in learning and memory? Neurobiol. Aging. 9:547548; 1988. Thompson, R. F.; Krupa, D. J.: Organization of memory traces in the mammalian brain. Annu. Rev. Neurosci. 17:519549; 1994. WoodruffPak, D. S.: Aging and classical conditioning: Parallel studies in rabbits and humans. Neurobiol. Aging 9:511522; 1988. WoodruffPak, D. S.; Thompson, R. F.: Classical conditioning of the eyelid response in rabbits as a model system for the study of brain mechanisms of learning and memory in aging. Exp. Aging Res. 11:109122; 1985. WoodruffPak, D. S.; Papka, M.; Ivry, R. B.: Cerebellar involvement in eyeblink classical conditioning in humans. Neuropsychology 10:443458; 1996. WoodruffPak, D. S.; Logan, C. G.; Thompson, R. F.: Neurobiological substrates of classical conditioning across the life span. Ann. NY Acad. Sci. 608:150178; 1990. WoodruffPak, D. S.; Papka, M.: Alzheimers disease and eyeblink conditioning: 750 ms trace vs. 400 ms delay paradigm. Neurobiol. Aging 17:397404; 1996. WoodruffPak, D. S.; Finkbiner, R. G.; Sasse, D. K.: Eyeblink conditioning discriminates Alzheimers patients from nondemented aged. Neuroreport 1:4549; 1990. WoodruffPak, D. S.; Papka, M.; Simon, E. W.: Eyeblink classical conditioning in Downs Syndrome, Fragile X syndrome, and normal adults over and under age 35. Neuropsychology 8:1424; 1994.

21. 22.

23.

24.

25. 26. 27.

28.

29. 30. 31. 32.

33. 34. 35. 36. 37.