• anticipation;
  • cerebral blood flow;
  • classical conditioning;
  • gastrointestinal motility;
  • rectosigmoid colon;
  • transcutaneus electrical nerve stimulation


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Abstract  The relationship between the central processes of classical conditioning and conditioned responses of the gastrointestinal function is incompletely understood in humans. We tested the hypothesis that the rectosigmoid motility becomes conditioned with anticipatory painful somatosensory stimulus and that characteristic brain areas become activated during anticipation. In nine right-handed healthy male subjects, a loud buzzer (CS, conditional stimulus) was paired with painful transcutaneus electrical nerve stimulation to the right hand (unconditional stimulus). Rectosigmoid muscle tone measured by the barostat as the intrabag volume, phasic contractions of the bowel measured as the number of phasic volume events (PVEs), and regional cerebral blood flow assessed by positron emission tomography (PET), were measured before and after conditioning. Following conditional trials, the bag volume after CS alone did not show significant changes between before and after the stimulus, but the number of PVEs after 2-minute interval of the CS alone was significantly greater than that before the stimulus (P < 0.05). The PET data showed the conditioning elicited significant cerebral activation of the prefrontal, anterior cingulate, parietal and insula cortices (P ≤ 0.001, uncorrected). Rectosigmoid motility can be conditioned with increase in phasic contractions in humans.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

In classical or Pavlovian conditioning, the conditional stimulus (CS), which is a neutral stimulus paired with an uncomfortable unconditional stimulus (US) previously, comes to elicit behavioural and physiological responses and the US alone.1–3 This learning process provides a model to understand anticipatory reports of pain and anticipatory gastrointestinal symptoms in situations that are not objectively threatening or painful.4

Little is known about the process of anticipatory response in gastrointestinal motility in humans. Physical and psychological actual stress induces significant changes in gastrointestinal motility, which includes smooth muscle tone and phasic contractions of the gastrointestinal tract.5 Patients with irritable bowel syndrome (IBS) show greater responses with abnormal patterns in the duodenal and colonic motility than healthy subjects during stress.6 The studies of the Pavlovian conditioning paradigm in the animal model revealed that the anticipatory stimulus elicited the same gastrointestinal responses as a delivered actual stimulus.7 In this model, the CS caused a significant increase in colonic spike burst frequency compared to basal values after repeated foot shock.7 Moreover, epidemiological studies revealed that post-traumatic stress disorder (PTSD)8 and a history of sexual or physical abuse,9 which tend to be accompanied with anticipatory fear/anxiety,10 had a high prevalence in patients with IBS. These phenomena suggest that central enhancement induced by associative learning may affect changes in gastrointestinal function.11

Classical conditioning is considered to be a model to understand anticipatory responses to aversive events, which is an essential component of how the brain–gut interaction develops in functional gastrointestinal disorders. Recently, central process of anticipatory responses has been investigated by several paradigm of brain imaging studies.12–15 In spite of research progress of brain imaging studies, there have been few studies to observe conditioned response in both brain and gastrointestinal motility function. It has been established that conditioned response can be observed by pairing a painful somatosensory stimulus with a neutral stimulus.16 In this study, we tested the following hypothesis: (i) the rectosigmoid motility becomes conditioned with increasing smooth muscle tone and increasing number of phasic contractions in humans and (ii) characteristic brain areas become activated during anticipation regardless of the stimulus intensity.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References


Nine right-handed healthy male subjects (mean age 24 ± 1 years; 19–29 years) were recruited from Tohoku University Campus in Sendai, Japan. All participants were free of gastrointestinal complaints and had not taken any medications within 4 weeks prior to testing. Each participant in this study underwent a medical history evaluation and was given a physical examination. Written informed consent was obtained from all participants, and this study was approved by the Tohoku University Ethics Committee.

Measurement of rectosigmoid function

The experiment was performed after a fasting period of at least 9 h. The subjects were placed in supine position and were instructed not to move during each session because of positron emission tomography (PET) scanning at the same time. A computer-driven barostat (Synectics Visceral Stimulator; Synectics, Stockholm, Sweden) was used to assess the rectosigmoid function.17–19 A polyethylene bag (diameter, 9 cm; length, 9 cm; volume, 0–500 mL), which was tightly fixed at both ends to a catheter, was inserted into the rectosigmoid colon of each subject and placed with distal end of the bag 10 cm from the anal verge 30 min before the study. The biomechanical properties of the bag were determined by pressure–volume measurements with the bag outside of a subject (ex vivo; Fig. 1). At volumes of less than 430 mL, the bag itself did not contribute to resistance to inflation.


Figure 1. Barostat bag compliance measured ex vivo. The pressure–volume curve demonstrated operation in the low-elastance portion for operating volumes <430 mL. In this range, the bag itself did not contribute to resistance to inflation, assuring that barostat measurements reflect the mechanical characteristics of the surrounding tissue.

Download figure to PowerPoint

Before the protocol, an initial distension in which the balloon pressure was increased from 0 to 40 mm Hg in 2 mm Hg steps at 10-second intervals was performed to reduce variability in compliance and to confirm the reproducibility. Thereafter rectal compliance was assessed by graded inflation until the first painful sensation or a maximal pressure of 40 mm Hg in the same way of the initial distension. During the protocol, the intra-operating pressure of the barostat bag was kept constant at 10 mm Hg as one of the standard methods for measuring colonic tone.19 On the other hand, there is the other standard method to consider the minimal distending pressure as the intra-operating pressure.17–19 However, no subjects in the present study showed that each minimal distending pressure (median 8 mm Hg; 6–8 mm Hg) exceeded the operating pressure. Besides, the operating pressure was much lower compared with the threshold of the first painful sensation (median 30 mm Hg; 22 to >40 mm Hg), which showed reproducibility in each subject.

Measurement of brain activation

Using a similar technique, which we have described in the previous report,20 regional cerebral blood flow (rCBF) was measured. Subjects were instructed to lie on their back in the PET scanner and to minimize head movement and keep their eyes closed during the scanning (for 70 s). Using a 68Ge/68Ga radiation source, transmission scans were carried out prior to PET scanning. [15O]-labelled water (Tohoku University Cyclotron Radioisotope Center) was injected into the right arm vein 10 s before the beginning of each stimulus session. Ten seconds later, the radioactivity in the brain reached a plateau and an increase in rCBF was detected by the PET scanning as an index of neural activity evoked by the stimulus. As shown in Fig. 2, five scans of rCBF in each subject were measured using PET scanner in three-dimensional sampling mode (HEADTOME V SET-2400W, Shimadzu, Kyoto, Japan).21 The scanner produced 63 horizontal slices with a separation of 3.125 mm, an axial field of view of 200 mm, an in-plane resolution of 590 mm, a full width at half maximum (FWHM) and an axial resolution of 3.9 mm FWHM. To ensure that radioactivity levels in each subject returned to baseline before starting a new scan, a 10-minute interval was given between successive scans.


Figure 2. The protocol in this study. Simple tones of buzzer horn were used as conditional stimulus (CS) and following transcutaneus electrical stimulation (TENS) to the right hand were used as unconditional stimulus (US). Only the CS tones were administrated as a preconditioning trial. After three 70-second sequences of conditional trials in which the CS was paired with the US, three additional test sequences were presented in random order; they consisted of the CS presented alone, CS paired with 7-mA US, or CS paired with 4-mA US as postconditioning trials. Subjects were exposed seven times to a loud buzzer in each trial. The US was started just after each tone was finished (no overlap). PET scanning was performed at the resting period as a background, and the pre- and postconditioning trials.

Download figure to PowerPoint


The protocol for the present study is shown in Fig. 2. There were three sessions; preconditioning, conditional and postconditioning trials. Subjects were exposed seven times to a loud buzzer (500 Hz with an intensity of 87 dB) lasting 1 s and being followed by a 9 s break. This sequence served as the CS. For the first sequence, only the CS tones were administrated as a preconditioning trial.

The US, which followed the CS during the conditional trials and a part of the postconditioning trials, was composed of transcutaneous electrical nerve stimulations (TENS; OG GIKEN AUDIO TREATER EF-501, Okayama, Japan) delivered to the back of the right hand at a frequency of 15 Hz with two different levels of intensity (7 or 4 mA). The US started just after each tone was finished and the stimulus period lasted 70 s (Fig. 2). After three sets of the CS or the postconditioning CS sequence, high-mA TENS was applied as the US. After the postconditioning CS sequence, low-mA TENS was applied as weak US. After the pre- or postconditioning CS-alone sequence, the US was not applied. In the postconditioning session, stimulus intensities of 0 (sham), 4 and 7 mA were given in random order.

The PET scanning was performed at the resting period as a background, and the pre- and postconditioning trials for each subject (five injections per scans, see Fig. 2). Each combination of the stimulus (the CS with/without the US) with break (10-second duration) was repeated seven times because the PET technique requires a 70-second recording window for each scan. The intra-bag pressure of barostat was kept at 10 mm Hg to measure changes in the bag volume in the rectosigmoid colon.

The subjects were also asked in the scanner to verbally rate the intensity of overall anxiety on a 0–10 point scale, with 0 representing no anxiety and 10 being the most anxious, before and after all series of the sessions.


The intrabag volume in the rectosigmoid colon was measured continuously and its variations were visually analysed. Mean bag volume over each 2-minute interval served as a measure of muscle tone, and number of phasic volume events (PVEs), served as a measure of phasic contractions according to the reported standard method.17,18 In the present study, 2-minute interval for the analysis of barostat measurement was selected not to fail to observe changes in the rapid volume waves.17 To control for occasional, minor changes in colorectal tone, the volume had to differ more than 10% from the baseline tone occurring at a frequency of 1–4 min−1 to be characterized as a change17 (see Fig. 3). Movement artifacts were defined as sudden changes in bag volume that did not continue for more than 15 s and/or did not differ more than 10% from baseline;17 these artifacts were excluded from data analysis. Changes in the bag volume or number of PVEs from each 2-minute baseline interval just before the stimulus (baseline interval) to each 2-minute interval just after the beginning of the stimulus (stimulus interval), and each following 2-minute interval (poststimulus interval), were considered to represent the colorectal wall reactivity to the CS with/without the US (Fig. 3). The paired Student's t-test or Wilcoxon's rank-sum test was used for comparing the rectosigmoid function in the 2-minute baseline, stimulus and poststimulus intervals of each trial. Alpha level was set at 5% for these statistical analyses.


Figure 3. Examples of tracings of the barostat. Changes in the rectosigmoid bag volume were measured during the preconditioning trial (top; A), the postconditioning CS-alone trial (middle; B) and the postconditioning CS/US trial (bottom; C) using the barostat. The tracings that were obtained from one of the healthy 26-year-old male subjects were shown as the intrabag pressure (at the top) and the intrabag volume (at the bottom), respectively. Phasic volume events (P) were considered to be minor changes in colorectal tone, which differed more than 10% from the baseline tone. Artifacts (A) were considered to be sharp waves, that were parenthetically observed and that did not continue for more than 15 s and/or did not differ more than 10% from baseline tone.

Download figure to PowerPoint

The PET data were transferred to a super computer (NEC SX-4/128H4, Tokyo, Japan) at the computer centre of Tohoku University through the optical network. The image reconstruction of all brain area was carried out using the three-dimensional filtered back projection algorithm.22 The PET image data were analysed using standard software (Statistical Parametric Mapping; SPM99, The Welcome Department of Cognitive Neurology, London) according to the method of Friston et al.23 All brain slices were analysed. The PET images were realigned, spatially normalized and transformed into an approximate Talairach–Tournoux stereotactic space, 3D Gaussian filtered (FWHM; 13 mm) and proportionally scaled to account for global confounders. The size of each voxel was set at 2 × 2 × 2 mm. A t-test was used to compare rCBF differences between the pre- and postconditioning CS-alone trials as a primal analysis for the effect of the conditioning. In addition, rCBF during the postconditioning CS with high- or low-mA TENS trial was compared with that during the preconditioning CS-alone trial as secondary analyses for the effect of both nociception and conditioning. As an additional analysis, brain regions manifesting linear correlations to the mean bag volumes or the number of PVEs on the barostat measurement were also examined using simple regression analysis in SPM99. According to the reported methods of the 3D brain imaging studies,22,23 we set alpha equal to 0.1% (uncorrected for multiple comparisons) as the region of significant differences. The region, which showed the significant activity correlations, was identified on the basis of Talairach co-ordinates.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

No subjects had a history of functional/organic gastrointestinal disorders, psychiatric/psychological disorders or physical/sexual abuse. No abnormality was found on physical examination including failure to anal relaxation with a rectal digital examination in each subject. All subjects completed the full protocol. All the subjects reported pain to the right hand and different given stimulus intensities during the postconditioning buzzer (CS) with high- or low-mA stimulus (US) trials. They did not report any pain or discomfort to the right hand in the buzzer-alone test trials. The buzzer with TENS or the buzzer alone did not induce any gastrointestinal symptoms. The anxiety scores did not show a significant change between before and after the protocol [median 4 (0–10) vs 2 (0–10), P > 0.1].

Assessment of rectosigmoid function

The rectosigmoid pressure and volume measurements were recorded at each distension step in the bag were plotted as mean values for the subjects in Fig. 3. The mean bag volume during 2-minute baseline interval was not significantly different among the sessions before and after the conditioning (Table 1). Example of actual barostat traces during the pre/postconditioning CS-alone (buzzer-alone) trials and the postconditioning CS with the US (buzzer with TENS) trials are shown in Fig. 4. In the postconditioning CS + high-mA US trial, the mean bag volume during 2-minute poststimulus interval was significantly smaller than that during 2-minute baseline interval (P < 0.05, Table 1). In the preconditioning trial and the postconditioning CS-alone and CS + low-mA US trials, the mean bag volume during the stimulus or poststimulus intervals did not show significant difference compared to that during each baseline interval. Thus, no conditioned effect was demonstrated for rectosigmoid muscle tone.

Table 1.  Mean bag volume and number of PVEs in the rectosigmoid colon during pre- and postconditioning trials
 Mean bag volume (mL)Number of PVEs (min−1)
Baseline intervalStimulus intervalPoststimulus intervalBaseline intervalStimulus intervalPoststimulus interval
  1. CS, conditional stimulus; US, unconditional stimulus; PVEs, phasic volume events.

  2. The duration of each interval is 2 min. Data were shown as mean ± SE or median with range. *P < 0.05 vs baseline interval (Student's t-test or Wilcoxon's rank-sum test).

CS alone/preconditioning43 ± 841 ± 840 ± 80 (0–1.5)0.5 (0–1.5)0.5 (0–1.5)
CS alone/postconditioning36 ± 1136 ± 1134 ± 130 (0–2)0.5 (0–3)  1 (0–2.5)*
CS + US (4 mA)/postconditioning48 ± 2044 ± 1538 ± 110.5 (0–1.5)1 (0.5–2)  1 (0.5–2)*
CS + US (7 mA)/postconditioning65 ± 2963 ± 3047 ± 18*0 (0–1.5)1 (0–3)  1 (0–2.5)*

Figure 4. Pressure–volume curve reflecting the rectal compliance during intermittent isobaric distension in the subjects. The curve represents the mean value of the barostat bag volume (±SE) for each step of distension.

Download figure to PowerPoint

In the postconditioning CS-alone trial, the number of PVEs during the 2-minute poststimulus interval was significantly greater than that during the immediately preceding 2-minute baseline interval (P < 0.05, Table 1). The number of PVEs during the poststimulus intervals were significantly greater than those during the baseline intervals in the postconditioning CS + low-mA US (P < 0.05) and CS + high-mA US (P < 0.05) trials, respectively. There were no significant differences in the number of PVEs in the preconditioning trial (Table 1). These data support a conditioning effect for colonic phasic contractions.

Assessment of central activation

The average PET data from all the subjects showed the conditioning elicited significant activation of the left lateral prefrontal, right anterior cingulate, bilateral parietal cortices, right insula, right pons and left cerebellum (P ≤ 0.001, uncorrected; Table 2 and Fig. 5) when comparing rCBF differences between pre- and postconditioning CS-alone trials of PET images. Comparing the postconditioning CS with 7-mA US with the preconditioning CS-alone trials, there was significant more activation of the bilateral primary sensory, left frontal, temporal, posterior cingulate and occipital cortices and bilateral pons (P ≤ 0.001, uncorrected; Table 3). Comparing the postconditioning CS with 4-mA US with the preconditioning CS-alone trials, there was significant more activation of the left primary sensory, prefrontal, anterior cingulate, parietal and primary motor cortices, bilateral precentral gyrus, left putamen and left pons (P ≤ 0.001, uncorrected; Table 4).

Table 2.  Areas of rCBF significantly increased in the postconditioning CS-alone trial compared to the preconditioning CS-alone trial (P ≤ 0.001, uncorrected)
Area (Broadmann area)HemisphereTalairach co-ordinateZ-score
  1. rCBF, regional cerebral blood flow.

  2. *Two different activated areas in the right pons were discriminated.

Prefrontal cortex (46)Left−4240104.1
Anterior cingulate cortex (32)Right1434303.4
Insula (13)Right362−43.8
Parietal cortex (40)Left−34−32403.6

Figure 5. Conditioning effects on regional cerebral blood flow (rCBF). Parametric maps of regional cerebral blood flow increase during the postconditioning CS-alone trial compared with the preconditioning CS-alone trial is superimposed on Talairach–Tournoux stereotactic space. Sagittal, coronal and axial views are centred at 12, 34, 30, in the right anterior cingulate cortex (left; A), at −42, 40 and 10, in the left prefrontal cortex (middle; B) and at 36, 2, −4, in the right insula (right; C), respectively. In the view of A, activation of the right parietal cortex also can be seen. Significant changes are marked with a split-grey scale (P ≤ 0.001, uncorrected).

Download figure to PowerPoint

Table 3.  Areas of rCBF significantly increased in the CS with full intensity US trial compared to the preconditioning CS-alone trial (P ≤ 0.001, uncorrected)
Area (Broadmann area)HemisphereTalairach co-ordinateZ-score
  1. rCBF, regional cerebral blood flow.

Primary sensory cortex (1)Left−34−28563.8
Primary sensory cortex (2)Left−48−22363.4
Orbitofrontal cortex (11)Left−3436−164.1
Precentral gyrus (6)Left−382263.9
Posterior cingurate cortex (31)Left−18−52203.6
Middle temporal gyrus (22)Left−42−4263.8
Occipital cortex (18)Left−20−82244.0
Table 4.  Areas of rCBF significantly increased in the CS with weak intensity US trial compared to the preconditioning CS-alone trial (P ≤ 0.001, uncorrected)
Area (Broadmann area)HemisphereTalairach co-ordinateZ-score
  1. rCBF, regional cerebral blood flow.

Primary somatosensory cortex (2)Left−34−32403.4
Prefrontal cortex (9)Left−408303.8
Prefrontal cortex (10)Left−3640203.8
Orbitofrontal cortex (11)Left−3434−163.6
Anterior cingulate cortex (24)Left−410364.6
Insura (13)Left−42−1643.7
Parietal cortex (40)Left−50−30463.9
Superior frontal gyrus (6)Left−66583.7
Middle frontal gyrus (6)Left−10−6503.7
Precentral gyrus (6)Left−560203.7
Primary motor cortex (4)Left−34−24603.4

For the postconditioning CS-alone trial, there were no significant correlations between rCBF in any region and the mean bag volumes or the number of PVEs.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

In the present study, the loud buzzer used prior to conditioning as a conditioned stimulus (CS) did not cause any alteration in rectosigmoid motility. However, following a series of conditional trials in which the buzzer was paired with painful electrical stimulation to the right hand, the buzzer-alone elicited increases in the phasic contractions of the rectosigmoid colon, which were similar to those seen following the conditioned stimulus plus the US. This provides evidence for Pavlovian conditioning of phasic motor responses. However, we did not find evidence for conditioning of the tonic motor response (barostat volume) or subjective pain; following conditional trials, the CS alone did not elicit changes in barostat volumes or reports of any gastrointestinal symptoms in the healthy subjects.

Little has been known about the anticipatory motor response in digestive system in humans. Previously, Naliboff et al.15 and our group20 reported that healthy human subjects reported slight unpleasantness/pain in response to sham distention of the colon after actual painful distention, and Mertz et al.24 reported the same for sham distention of the stomach in patients with functional dyspepsia. However, these reports have not investigated gastrointestinal motility function. Only the animal study with rat model revealed that conditioned fear after repeated foot shock as US caused a significant increase in colonic spike burst frequency but failed to affect jejunal motility.7 The colonic motility changes that have been identified in the present study are considered to be as one of secondary phenomena that would occur during the anticipation of pain. Anticipation affects the autonomic nervous function. The observation that anticipation of painful/aversive stimulus resulted in changes in heart rate was confirmed by several studies in both human and animal models.25–27 Tests with drugs blocking the sympathetic or parasympathetic fibres revealed that the conditioned group, showing anticipatory fear, actually had a large sympathetic increase that was partly masked by a simultaneous parasympathetic increase.28,29 Considering the previous studies on autonomic nervous activity during anticipation, increase in colorectal motility observed following a series of painful stimuli to the hand may result as dominant parasympathetic arousal. Further pharmacological intervention should be needed to confirm this hypothesis.

On the other hand, the anticipatory rectosigmoid motor responses could represent nonspecific arousal or anxiety. It has been reported that psychological stress induces significant changes in gastrointestinal motility, which were associated with alterations in autonomic nervous activity.30,31 However, nonspecific arousal or anxiety cannot explain main effects of the conditioned phasic contractions of the rectosigmoid colon observed in this study for two reasons: (i) the overall anxiety score which is considered to be very global did not show a significant difference between before and after the series of painful stimuli. (ii) The increases in PVEs were limited to the poststimulus interval and were not seen during the baseline interval in the postconditioning CS-alone trial although this baseline interval was preceded by a series of painful stimuli.

Considering the conditioning effect in the brain, our findings of the brain imaging (Table 2 and Fig. 5) were in accordance with previous studies showing cerebral activation in the frontal and parietal cortices following Pavlovian conditioning.12–14 Activation of the prefrontal cortex was seen during somatic stimulus, and has been implicated in cognitive appraisal of the stimulus.32 In addition, significant cortical activation in the anterior cingulate cortex (ACC), which is believed to play a role in mediating the affective qualities of the pain experience33,34 and expectation of pain,35 and in the insula which serves as limbic integration cortex36 was also seen as anticipatory responses in this study. Therefore, our results support that activation of the cognitive- and affective-related brain regions may contribute to the learned anticipatory responses and that this learned process was confirmed after the conditional trials in this experimental model. However, this model has been set up to study the anticipatory colonic motor response and brain activation patterns that have been identified only reflected that. The direct relationships between the brain activation and the gastrointestinal response during anticipation have not been clarified with this model.

When comparing brain-imaging data between the postconditioning buzzer with high- or low-mA stimulus and the preconditioning buzzer-alone trials, increased rCBF not only in the left primary somatosensory cortex but also in the prefrontal cortex was observed (Tables 3 and 4). These findings were in concordance with the previous reports on the central processing of painful stimulus using with brain-imaging studies,32,35,37 Thus, in the present study, these comparisons revealed that the transcutaneus electrical stimulation to the right hand induced painful sensation. Furthermore, activation of the cognitive- and affective-related brain regions was observed in common, suggesting that cerebral responses involved in painful/fearful anticipation might be different from the nociceptive process.38

There are some limitations in this study. First of all, participants of the study were limited to small group of healthy volunteers from a local university. Lee et al. have investigated differences in gastrointestinal symptom severity in males vs females and variations with menstrual cycles.39 Female subjects might show different patterns in the brain and/or colonic motility function for the conditioned responses. Secondly, additional examinations such as an anorectal morphological study were not explored. However, existence of severe megarectum40 and/or the other anorectal disorders,41 which might affect rectal wall motor function, were unlikely because no subjects had been reported any problems of bowel movement and any abnormal physical findings. Finally, the reliable visceral sensory tests (e.g. the ascending method of limits and/or the random sequence)19 were not used because of the limited time in the PET scanning room. Despite of these limitations, we believe that this study could be worth to report the conditioned phenomena in this model as a first step to understand anticipatory colonic motility responses to somatosensory painful stimulus in humans. The available data on conditioned gastrointestinal responses are very limited and deserve further studies.

In summary, the Pavlovian conditioning study is significant because of positive findings that the conditioned phenomenon in this model is a first step to understand the anticipatory colonic motility responses. Significant increases in colonic phasic contractions and significant increases in cerebral blood flow in the cognitive- and affective-related cortical regions were observed in this study. This conditioning paradigm could be a model to investigate anticipatory responses in gastrointestinal motility and brain function, which may contribute to development of functional gastrointestinal disorders. We concluded that the rectosigmoid motility could become conditioned by pairing a painful somatosensory stimulus with a neutral stimulus in humans.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This study was supported by Grant-in-Aid for Scientific Research Nos 833-5020-11680778, 622-5637-13897021 and 7201-3400-15390218 from the Ministry of Education, Science and Culture of Japan and Grant-in-Aid for Scientific Research Nos H11-Nou-003, H13-Chouju-028 and H14-Si-9 from the Ministry of Health, Welfare and Labour of Japan (PI: Shin Fukudo).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Rescorla RA. Behavioral studies of Pavlovian conditioning. Ann Rev Neurosci 1988; 11: 32952.
  • 2
    Bjorkstrand PA. Effects of conditioned stimulus pre-exposure on human electrodermal conditioning to fear-relevant and fear-irrelevant stimuli. Biol Psychol 1990; 30: 3550.
  • 3
    Dadds MR, Bovbjerg DH, Redd WH, Cutmore TR. Imagery in human classical conditioning. Psychol Bull 1997; 122: 89103.
  • 4
    Cameron OG. Interception: the inside story—a model for psychosomatic processes. Psychosom Med 2001; 63: 697710.
  • 5
    Rao SS, Hatfield RA, Suls JM, Chamberlain MJ. Psychological and physical stress induce differential effects on human colonic motility. Am J Gastroenterol 1998; 93: 98590.
    Direct Link:
  • 6
    Fukudo S, Nomura T, Muranaka M, Taguchi F. Brain-gut response to stress and cholinergic stimulation in irritable bowel syndrome. A preliminary study. J Clin Gastroenterol 1993; 17: 13341.
  • 7
    Stam R, Croiset G, Akkermans LM, Wiegant VM. Effects of novelty and conditioned fear on small intestinal and colonic motility and behaviour in the rat. Physiol Behav 1995; 58: 8039.
  • 8
    Irwin C, Falsetti SA, Lydiard RB et al. Comorbidity of posttraumatic stress disorder and irritable bowel syndrome. Clin Psychiatry 1996; 57: 5768.
  • 9
    Drossman DA, Leserman J, Nachman G et al. Sexual and physical abuse in women with functional or organic gastrointestinal disorders. Ann Intern Med 1990; 113: 82833.
  • 10
    Bremner JD, Vythilingam M, Vermetten E et al. Cortisol response to a cognitive stress challenge in posttraumatic stress disorder (PTSD) related to childhood abuse. Psychoneuroendocrinology 2003; 28: 73350.
  • 11
    Stam R, Akkermans LM, Wiegant VM. Trauma and the gut: interactions between stressful experience and intestinal function. Gut 1997; 40: 7049.
  • 12
    Fredrikson M, Wik G, Fischer H, Andersson J. Affective and attentive neural networks in humans: a PET study of Pavlovian conditioning. Neuroreport 1995; 7: 97101.
  • 13
    Hugdahl K, Berardi A, Thompson WL et al. Brain mechanisms in human classical conditioning: a PET blood flow study. Neuroreport 1995; 6: 17238.
  • 14
    Fischer H, Andersson JL, Furmark T, Fredrikson M. Fear conditioning and brain activity: a positron emission tomography study in humans. Behav Neurosci 2000; 114: 67180.
  • 15
    Naliboff BD, Derbyshire SW, Munakata J et al. Cerebral activation in patients with irritable bowel syndrome and control subjects during rectosigmoid stimulation. Psychosom Med 2001; 63: 36575.
  • 16
    Kandel ER, Kupfermann I, Iversen S. Learning and memory. In: KandelER, ScwartzJH, JessellTM, eds. Principles of Neural Science, 4th edn. New York: McGraw-Hill, 2000: 122747.
  • 17
    Bell AM, Pemberton JH, Hanson RB, Zinsmeister AR. Variations in muscle tone of the human rectum: recordings with an electromechanical barostat. Am J Physiol 1991; 260: G1725.
  • 18
    von der Ohe, Hanson RB, Camilleri M. Comparison of stimultaneous recordings of human colonic constructions by manometry and a barostat. Neurogastroenterol Motil 1994; 6: 21322.
  • 19
    Whitehead WE, Delvaux M. Standardization of barostat procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. Dig Dis Sci 1997; 42: 22341.
  • 20
    Hamaguchi T, Kano M, Rikimaru H et al. Brain activity during distention of the descending colon in humans. Neurogastroenterol Motil 2004; 16: 299309.
  • 21
    Fujiwara T, Watanuki S, Yamamoto S et al. Performance evaluation of a large axial field-of-view PET scanner: SET-2400W. Ann Nucl Med 1997; 11: 30713.
  • 22
    Cherry SR, Dahlbom M, Hoffman EJ. Evaluation of a 3D reconstruction algorithm for multi-slice PET scanners. Phys Med Biol 1992; 37: 77990.
  • 23
    Friston KJ, Frith CD, Liddle PF, Frackowiak RS. Comparing functional (PET) images: the assessment of significant change. J Cereb Blood Flow Metab 1991; 11: 6909.
  • 24
    Mertz H, Fullerton S, Naliboff B, Mayer EA. Symptoms and visceral perception in severe functional and organic dyspepsia. Gut 1998; 42: 81422.
  • 25
    Obrist PA, Wood DM, Perez-Reyes M. Heart rate during conditioning in humans: effects of UCS intensity, vagal blockade, and adrenergic block of vasomotor activity. J Exp Psychol 1965; 70: 3242.
  • 26
    Santibanez-H G, Schroeder H. Does heart rate differentiate neurotic from normal people in conditioned reflex test? Pavlov J Biol Sci 1988; 23: 1521.
  • 27
    Kalin NH, Shelton SE, Davidson RJ et al. A new method for aversive Pavlovian conditioning of heart rate in rhesus monkey. Physiol Behav 1996; 60: 10436.
  • 28
    Iwata J, LeDoux JE. Dissociation of associative and nonassociative concomitants of classical fear conditioning in the freely behaving rat. Behav Neurosci 1988; 102: 6676.
  • 29
    Nijsen MJ, Croiset G, Diamant M et al. Conditioned fear-induced tachycardia in the rat: vagal involvement. Eur J Pharmacol 1998; 350: 21122.
  • 30
    Fukudo S, Suzuki J. Colonic motility, autonomic function, and gastrointestinal hormones under psychological stress on irritable bowel syndrome. Tohoku J Exp Med 1987; 151: 37385.
  • 31
    Gue M, Junien JL, Bueno L. Conditioned emotional response in rats enhances colonic motility through the central release of corticotropin-releasing factor. Gastroenterology 1991; 100: 96470.
  • 32
    Derbyshire SW, Jones AK, Gyulai F et al. Pain processing during three levels of noxious stimulation produces differential patterns of central activity. Pain 1997; 73: 43145.
  • 33
    Bantick SJ, Wise RG, Ploghaus A et al. Imaging how attention modulates pain in humans using functional MRI. Brain 2002; 125: 31019.
  • 34
    Bernstein CN, Frankenstein UN, Rawsthorne P et al. Cortical mapping of visceral pain in patients with GI disorders using functional magnetic resonance imaging. Am J Gastroenterol 2002; 97: 31927.
    Direct Link:
  • 35
    Sawamoto N, Honda M, Okada T et al. Expectation of pain enhances responses to nonpainful somatosensory stimulation in the anterior cingulate cortex and parietal operculum/posterior insula: an event-related functional magnetic resonance imaging study. J Neurosci 2000; 20: 743845.
  • 36
    Augustine JR. Circuitry and functional aspects of the insular lobe in primates including humans. Brain Res Rev 1996; 22: 22944.
  • 37
    Porro CA, Cettolo V, Francescato MP, Baraldi P. Temporal and intensity coding of pain in human cortex. J Neurophysiol 1998; 80: 331220.
  • 38
    Ploghaus A, Tracey I, Gati JS et al. Dissociating pain from its anticipation in the human brain. Science 1999; 284: 197981.
  • 39
    Lee OY, Mayer EA, Schmulson M, Chang L, Naliboff B. Gender-related differences in IBS symptoms. Am J Gastroenterol 2001; 96: 218493.
    Direct Link:
  • 40
    Chiarioni G, Bassotti G, Germani U et al. Idiopathic megarectum in adults. An assessment of manometric and radiological variables. Dig Dis Sci 1995; 40: 228692.
  • 41
    Felt-Bersma RJ, Sloots CE, Poen AC et al. Rectal compliance as a routine measurement: extreme volumes have direct clinical imp act and normal volumes exclude rectum as a problem. Dis Colon Rectum 2000; 43: 17328.