Digestive Diseases and Sciences

, Volume 55, Issue 2, pp 384–391

Effect of Meal Ingestion on Ileocolonic and Colonic Transit in Health and Irritable Bowel Syndrome


  • Annemie Deiteren
    • Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), College of MedicineMayo Clinic
    • Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), College of MedicineMayo Clinic
  • Duane Burton
    • Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), College of MedicineMayo Clinic
  • Sanna McKinzie
    • Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), College of MedicineMayo Clinic
  • Archana Rao
    • Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), College of MedicineMayo Clinic
  • Alan R. Zinsmeister
    • Division of Biomedical Statistics and Informatics, Department of Health Sciences Research, College of MedicineMayo Clinic
Original Article

DOI: 10.1007/s10620-009-1041-8

Cite this article as:
Deiteren, A., Camilleri, M., Burton, D. et al. Dig Dis Sci (2010) 55: 384. doi:10.1007/s10620-009-1041-8



Postprandial symptoms in irritable bowel syndrome (IBS) have been associated with increased bowel contractility.


To compare ileocolonic and colonic responses to feeding in health and IBS.


We prospectively analyzed data from separate research trials in 122 IBS patients and 41 healthy volunteers. Ileocolonic transit (ICT) was evaluated before (colonic filling [CF]3h) and immediately after (CF4h) a standard lunch at 3 h 45 min, and 2 h thereafter. The colonic geometric center (GC) was calculated 2 h (GC6h) after lunch ingested at 4 h (GC4h) and directly after (GC8h) a standard dinner ingested at 7 h 45 min.


ICT immediately after eating was higher in IBS diarrhea predominant (IBS-D) patients than in the healthy cohort (23.1 ± 2.4 vs. 17.5 ± 2.8%, P = 0.059). ICT 2 h after lunch was similar between groups (P = 0.55). There was significant overall group differences in colonic transit 2 h post-lunch (P = 0.045), particularly in the IBS constipation predominant (IBS-C; GC6–GC4, Δ0.29 ± 0.08) patients versus healthy volunteers (Δ0.56 ± 0.12 GC units).


After feeding, ICT is increased in IBS-D, whereas colonic transit is blunted in IBS-C.


ColonicGastrocolic reflexGastroileal reflexIleocolonicResponse to food


Irritable bowel syndrome (IBS) is a common functional disorder in which abdominal pain or discomfort is associated with a change in bowel habits. The prevalence of IBS in the general population is ~10% [1]. The pathophysiological mechanisms are diverse. By Rome III consensus criteria, patients’ IBS phenotype can be classified according to the predominant bowel pattern: diarrhea predominant (IBS-D), constipation predominant (IBS-C), or mixed IBS (IBS-M).

Symptoms may occur or be exacerbated after a meal in patients with IBS [24], including abdominal pain, gas, abdominal distension, and urgency [5].

The colonic response to food, sometimes termed the “gastrocolic reflex” [6], results in an integrated increase in colonic tone [7] and phasic pressure activity [8] following meal ingestion. The mechanisms mediating this response involve cholinergic pathways [9, 10], as well as mediators such as serotonin [7, 11], gastrin [8], prostaglandin E1 [12], and cholecystokinin [8].

The first description of an impaired colonic response to food in patients with IBS is attributed to Connell et al. [13] in the 1960s. Using colonic manometry, they demonstrated that patients with IBS had an exaggerated increase in postprandial colonic activity. In a more recent study, Bouchoucha et al. [14] compared the colonic transit of radiopaque markers at rest and after eating a standard test meal; segmental differences in the patterns of colonic filling (CF) and emptying were observed in response to eating in healthy volunteers and patients with IBS. The generalizability of the study was limited by the sample size, the propensity for included IBS patients to have delayed colonic transit, and the lack of assessment of ileocolonic transit (ICT). ICT is characterized by bolus movements [15, 16] that coincide with simultaneous prolonged propagated contractions [17] and may result in abdominal discomfort.

The aim of this study was to quantify ileocolonic and colonic transit in response to meal ingestion in healthy volunteers and different subgroups of IBS patients.


Data Source

Data were retrieved from a database of prospectively performed motor, sensory, psychological, and autonomic function tests in 122 patients with IBS and 41 healthy volunteers [4]. Patients residing within 200 miles of the Mayo Clinic were recruited based on their primary presentation with IBS, identified by the Rome II criteria [18]. Volunteers were included as healthy controls if they had less than four positive gastrointestinal symptoms out of 19 on the Bowel Disease Questionnaire (BDQ [19]), scored as mild severity at worst or “sometimes” for greatest frequency.

Exclusion criteria included organic disease that might explain the patient’s symptoms and any structural or metabolic disease or condition that affects the gastrointestinal system, including diabetes. In addition, participants were ineligible if they had participated in another clinical trial within the preceding 30 days. The use of any IBS or anti-constipation medication within at least 7 days prior to gastrointestinal transit measurements was not permitted. Participants were allowed to continue low, stable doses of thyroid replacement, low-dose aspirin (81 mg/d), selective serotonin reuptake inhibitors (SSRIs), estrogen replacement and birth control pills, or depot estrogen injections.

All participants had provided written informed consent before participating in the research study. The current analysis was approved by the Mayo Clinic Institutional Review Board. Patients who had withdrawn authorization to use their records for future research purposes had their data removed from the analysis.

Scintigraphic Transit Studies

An adaptation of our established scintigraphic method was used to evaluate the ileocolonic and colonic response to food. After an overnight fast, a capsule containing 111In adsorbed on activated charcoal was administered. The methacrylate coating (Eudragit S-100, The Dow Chemical Company) of the capsule dissolves in a pH-sensitive manner upon reaching the alkaline terminal ileum, releasing the radioisotope into the lumen, thus, allowing the assessment of colonic transit. After the capsule had emptied from the stomach, documented by its relative position to a radioisotope marker placed on the right anterior iliac crest, participants ingested a 99mTc-labeled breakfast (315 kcal; two scrambled eggs, one slice of whole wheat bread, and one glass of whole milk). The 99mTc-sulfur colloid in the breakfast was used to estimate the CF at different time points as a surrogate for ileocecal transit.

Using a gamma camera, anterior and posterior abdominal scans of 2 min duration were acquired immediately following ingestion of the radiolabeled breakfast and at 3, 4, 6, and 8 h post-breakfast. Additional scans were completed at 24 and 48 h. An illustration of the experimental design is provided in Fig. 1.
Fig. 1

Experimental design

Three hours and 45 min after ingestion of the 99mTc-labeled breakfast, participants were served a standardized lunch of 462 kcal, which comprised one chicken breast, one potato, one pudding cup, three pads of butter, and 300 ml of water (19% protein, 43% carbohydrates, and 38% fat). The transfer of radioisotope from the ileum to the colon following meal ingestion was measured to assess the ileocolonic response to food. To quantify the magnitude of the colonic response to food, a standardized 607-kcal dinner was served 7 h and 45 min after the 99mTc-labeled breakfast. This dinner consisted of a roast beef sandwich, milk, and a cookie (19% protein, 47% carbohydrates, and 34% fat).

Data Analysis

99mTc counts were quantified within a 140-keV (±20%) window and 111In counts within a 247-keV (±20%) window. Count corrections were made for isotope decay, tissue attenuation, and downscatter of the 111In in the 99mTc window. A variable region of interest program was used to quantitate counts in each colonic segment.

The ileocolonic response to food, during and following lunch, was quantified by the transfer of 99mTc-labeled chyme from the ileum to the colon and was expressed by the difference in CF at different time points. CF, the proportion of the 99mTc-labeled breakfast that has accumulated in the colon, was measured before the 462-kcal chicken lunch (3 h), immediately after finishing lunch (4 h), and 2 h thereafter (6 h). The differences in CF were used to quantify the ICT in response to food: the immediate response was expressed as CF4h–CF3h and the delayed ileocolonic response as CF6h–CF4h.

The colonic response to food was estimated by the progression of 111In-labeled content through the colon, with respect to the different colonic segments: ascending colon (AC), transverse colon (TC), descending colon (DC), and rectosigmoid (RS), numbered as segments 1–4, respectively. Segment 5 signifies the expelled stool. Colonic response to food was expressed as the difference in geometric center (GC) before and after the roast beef dinner. The GC is the weighted average of the isotope distribution within the colon and is calculated by the sum of the products of the proportion of 111In counts in each colonic segment and the segment’s weighting factor:
$$ {\text{GC}} = [ (\% {\text{AC}} \times 1 ) + (\% {\text{TC}} \times 2) + (\% {\text{DC}} \times 3) + (\% {\text{RS}} \times 4) + (\% S \times 5)]/100 $$

The immediate colonic response to food was expressed as GC8h–GC6h, as the 8-h scan was acquired directly after finishing dinner. We assessed the delayed colonic response to food 2 h post-meal based on the chicken lunch by comparing GC6h to GC4h.

Statistical Analysis

Kruskal–Wallis tests were used compare IBS-D, IBS-C, and IBS-M patients and healthy controls. The Wilcoxon rank-sum test was used to compare IBS-D or IBS-C patients and healthy controls. The results are expressed as mean ± SEM (standard error of the mean).



The participants’ demographics are summarized in Table 1. One patient became ill during transit measurements and dropped out. Of the remaining 163 participants, 160 were female and three were male. All groups were similar in age, but body mass indices (BMIs) were higher in patients with IBS-D and IBS-M. The relationships of meal-related symptoms and participant subgroups are shown in Table 1.
Table 1

Participants’ characteristics






Number (males)

41 (0)

47 (0)

45 (3)

30 (0)

Age, years

34.0 ± 1.6

38.4 ± 1.5

35.3 ± 1.5

38.7 ± 2.3

BMI, kg/m2

25.1 ± 0.7

25.3 ± 0.5

28.8 ± 1.0

27.6 ± 0.8

Any meal-related symptoms, %





BMI body mass index

Ileocolonic Response to Food Ingestion

For healthy volunteers, the mean immediate ileocolonic response to food ingested at 3 h 45 min was 17.5 ± 2.8%; the delayed response 2 h post-meal was 29.3 ± 3.2% (Table 2, Fig. 2).
Table 2

Ileocolonic and colonic response to food








Immediate ileocolonic response


17.5 ± 2.8

19.7 ± 1.46

19.4 ± 2.4

23.1 ± 2.4

15.1 ± 2.8

Delayed ileocolonic response


29.3 ± 3.2

27.34 ± 1.57

29.5 ± 2.6

24.2 ± 2.3

28.4 ± 3.6

Immediate colonic response


0.16 ± 0.04

0.18 ± 0.02

0.18 ± 0.04

0.18 ± 0.03

0.20 ± 0.06

Delayed colonic response


0.56 ± 0.12

0.35 ± 0.05

0.29 ± 0.08

0.50 ± 0.09

0.26 ± 0.05

CF colonic filling, GC geometric center

Fig. 2

Box and whisker plots showing the immediate and later ileocolonic response to food as the difference in % of colonic filling (CF) between 3 and 4 h and 4 and 6 h, respectively. Note the trend towards an increased immediate ileocolonic response in IBS diarrhea predominant (IBS-D) patients compared to healthy volunteers

There was no statistical difference in the immediate ileocolonic response to food between groups (P = 0.075). However, ICT immediately following meal ingestion was higher in IBS-D patients and healthy volunteers (23.1 ± 2.4 vs. 17.5 ± 2.8%, P = 0.059, Fig. 3).
Fig. 3

Representative ileocolonic transit (ICT) scintiscans at 3, 4, and 6 h in a healthy volunteer (a) and a patient with IBS-D (b). A region of interest is drawn around the ascending colon. The intensity of the image reflects the concentration of counts in each region. Note that there is increased CF in the patient with IBS-D at the 4-h scan, reflecting increased ileocolonic transfer in response to the meal

There were no differences between subgroups in the delayed ileocolonic response 2 h post-meal (Table 2).

Colonic Response to Food Ingestion

The mean change in GC after the dinner at 7 h 45 min was 0.16 ± 0.04 GC units for healthy volunteers (Table 2, Fig. 4). This immediate postprandial change in GC was not significantly different between IBS patients and healthy volunteers. Two patients, one with IBS-D and one with IBS-M, showed net retrograde movement of colonic content after finishing the meal, as GC8h–GC6h was negative (Fig. 4).
Fig. 4

Box and whisker plots showing the immediate and later colonic response to food as the change in geometric center (GC) between 6 and 8 h and 4 and 6 h, respectively. Note the impaired colonic response 2 h post-lunch in IBS constipation predominant (IBS-C) patients

The group difference for the delayed colonic motor response 2 h post-lunch (GC6h–GC4h) was significant (P = 0.045); in particular, it was lower for IBS-C patients compared to healthy volunteers (0.29 ± 0.08 vs. 0.56 ± 0.12 GC units, respectively, Fig. 5).
Fig. 5

Representative colonic transit scintiscans at 4 and 6 h in a healthy volunteer (a) and a patient with IBS-C (b). Regions of interest are drawn around the ascending and transverse colon. The intensity of the image reflects the concentration of counts in each region. Note that there is less movement of radioisotope in the patient with IBS-C compared to the healthy volunteer, reflecting a blunted delayed colonic response to food in the chicken lunch


In this study, we assessed the ileocolonic and colonic transit responses to food ingestion in patients with IBS and healthy volunteers by measuring the progression of chyme through the ileocolonic regions following meal ingestions that occurred 4 and 8 h after the intake of a radiolabeled meal and a capsule that delivered activated charcoal particles to the ileocolonic junction. We demonstrated increased ICT in IBS-D and decreased colonic transit in IBS-C patients in response to meal ingestion.


The transfer of chyme from the terminal ileum to the colon occurs in bolus movements [15] that coincide with prolonged propagating ileal contractions [17, 20]. After meal ingestion, the number of boluses increases [21] with concurrent filling of the colon [15], suggesting a “gastroileal” reflex with propulsion of content [21]. This may be a neural reflex or it may result from augmented ileal flow as the residue of the first meal reaches the terminal ileum ~4 h after ingestion and its movement into the colon coincides with the second meal ingestion [22, 23]; alternatively, specific nutrients may activate ICT.

Neural reflexes may contribute to the ICT. The cholinergic mechanisms involved in increased ileocecal transit have been investigated in a canine model of the ileocolonic region in vivo, with evidence for the involvement of muscarinic M1 and M3 mechanisms [24, 25]. If symptoms are aggravated postprandially in IBS, these data suggest that further studies of selective antimuscarinic agents in patients with IBS are indicated.

Disturbances of ileocolonic transfer have been proposed as a pathophysiological mechanism for symptoms in IBS patients [26, 27]. We observed a numerical difference in ICT during meal ingestion in patients with IBS-D. Since there was no difference in small bowel transit between IBS subgroups (data published in detail in [4]), it seems unlikely that the numerical difference in ICT observed in the current study results from a difference in the arrival of residue from the breakfast meal.

The arrival of products of fat digestion or complex carbohydrate (such as short chain fatty acids [SCFA]) in the ileum may trigger ileal motility [15, 28] and ileocolonic transfers; the presence of fat concentrations associated with moderate steatorrhea can overwhelm this capacitance and result in diarrhea [29, 30]. This activation of ileal motility by SCFA is inhibited by the nonselective opiate blocker, naloxone, and calcium channel blocker. Our studies are limited by the relatively low fat content of the diet. However, Steed et al. [31] showed that, in normal subjects, equicaloric high- and low-fat meals did not induce different transit profiles. Further evaluation of higher concentrations of fat or SCFAs delivered to the ileocolonic region in IBS patients would be of significant interest.

Colonic Transit and Motility and Food Ingestion

We were not able to demonstrate a difference in the immediate colonic response following meal ingestion between IBS patients and healthy volunteers. Bazzocchi et al. [32] had reported the absence of an immediate colonic response following meal ingestion in patients with chronic functional diarrhea, in contrast to healthy volunteers. However, there are several differences in these studies. First, our scintigraphic measurement evaluates the unprepared colon, whereas Bazzocchi et al. performed colonic cleansing, which may have affected the residual content and its propulsion. Second, we used a 462-kcal lunch and a 607-kcal dinner (each consisting of ~20% protein, 45% carbohydrates, and 35% fat), whereas the other study used a high-fat, 1,000-kcal meal. The caloric content and meal composition may be important determinants of the magnitude of colonic response to food [8, 31, 3335]. Nevertheless, the meal calorie content in our studies was sufficient to stimulate colonic motility, since colonic motor responses to food have been documented with a calorie load comparable to ours [36] or lower (200 kcal [33]).

Electromyographic and manometric observations have demonstrated an increase in colonic tone and spike activity in IBS patients after eating [9, 14, 3739]. Studies of propulsion or colonic transit of radiopaque markers after meals [14] in healthy controls showed emptying of the cecum and ascending colon, and filling of the rectosigmoid. In IBS patients [14], eating resulted in emptying of the cecum-ascending colon, the left transverse colon, and the splenic flexure, without filling of the distal colon. Di Stefano et al. showed a normal tonic response to 200 kcal in healthy volunteers and to 400 kcal in IBS-C patients; in contrast, patients with IBS-D required 1,000 kcal to induce the normal increase in rectosigmoid tone [33]. The reduced colonic response to food reported by Bouchoucha et al. may reflect colonic function in patients with IBS-C [14]. The blunted colonic response to food in IBS-C patients 2 h postprandially in our study is consistent with the overall results of Bouchoucha et al. and with other reports of impaired colonic response to food in constipated patients [4044].

Colonic Response and Symptoms After Feeding

An abnormal colonic response to food may trigger postprandial symptoms in patients with IBS, though there is little objective evidence to date. Simrén et al. documented that the majority of patients with IBS consider their symptoms to be related to meals, particularly foods rich in carbohydrates and fat; they did not measure colonic transit in the same study [5]. We did not gather symptom data during the transit measurements [8]; however, all participants filled out bowel disease questionnaires prior to the transit studies, including an assessment of postprandial symptoms such as abdominal pain occurring immediately (0–30 min) or 30–120 min after a meal and the aggravation of symptoms after oral intake. These data, previously published in detail elsewhere [4], show that 95% of our healthy volunteers experience no postprandial symptoms, whereas ~70% of IBS patients report postprandial symptoms (Table 1). There was also a higher percentage of IBS-D patients who reported abdominal pain immediately following a meal. These data are consistent with the hypothesis that increased ileocolonic transfer in IBS-D patients might be associated with postprandial symptoms. However, this has to be tested prospectively with the measurement of motor responses and symptoms simultaneously. The lack of any difference in the immediate postprandial colonic response to food ingested at 4 and 8 h suggests that postprandial urgency may reflect distal colonic stimulation by feeding. This was not detectable in our study since the isotope was located in the proximal colon at 4–8 h.

Strengths and Limitations

One of the strengths of this study is the large number of participants included for the assessment of ileocolonic and colonic transit in response to eating in the different subgroups of IBS patients. If confirmed in prospective studies of transit and simultaneous symptom measurement, this may lead to better designed treatments that inhibit postprandial responses such as 5-HT3 antagonists [7, 45] and selective muscarinic antagonists [24, 25].

As the study was conducted retrospectively, there are some weaknesses; meals of a different caloric content were served at ~4 and ~8 h after the radiolabeled breakfast meal, and colonic response to food was assessed in response to both meals. The delayed colonic response to feeding was assessed in relation to the smaller lunchtime meal, as no scans were available at the 10-h time point (2 h after the larger meal).


In summary, this study demonstrates that scintigraphy can assess ileocolonic motor responses to food ingestion and suggests impaired colonic response to food in patients with IBS constipation predominant (IBS-C) and numerically increased ileocolonic response (P = 0.059) in patients with IBS diarrhea predominant (IBS-D). Future studies should explore the relationship of food ingestion to the pathophysiology of IBS and the relationship to symptoms measured simultaneously. These noninvasive studies may provide evidence for the colonic response to food as a potential target for pharmacological treatment to normalize the responses and relieve patients’ symptoms.


Dr. Camilleri’s work in IBS is supported in part by RO1 grant DK-54681 from the National Institutes of Health.

Copyright information

© Springer Science+Business Media, LLC 2009