Periodic colonic motor activity identified by 24-h pancolonic ambulatory manometry in humans


Address for correspondence Mr Devinder Kumar, Department of Surgery, St. George's Hospital, Blackshaw Road, Tooting, London SW17 0QT, UK. Tel.: + 44 181 7251301; fax: + 44 181 7253611.


Abstract The pattern of colonic motor activity in healthy humans has not been fully elucidated to date. The aim of this study was to evaluate colorectal motor activity employing 24-h ambulant pancolonic manometry. Ten healthy volunteers (6F, 4M), aged 19–31 years were studied. Motor activity was measured using two custom-made silicone coated catheters, each with five solid-state pressure transducers. No bowel preparation or sedation was used. The study period was 24 h. A total of 232 h of recording was obtained. Sixty-three high amplitude propagated contractions were observed, median six per 24-h period. Low-amplitude colonic contractile activity showed regional and diurnal variations. Frequency of contraction was highest in the right colon [median 5.26 cpm (cycles per minute)], and transverse colon and splenic flexure (median 5.15 cpm). The interval between colonic motor complexes was shortest in the transverse colon and splenic flexure. This study introduces a new technique for the evaluation of colorectal motor activity. Subjects were studied in an ambulant setting in their own environment ensuring that this method of study is as physiological as possible. This study demonstrates that colonic motor activity has two main components: high amplitude propagated contractions and low amplitude colonic contractile activity.


Colonic functional disorders account for a high proportion of the workload of gastrointestinal physicians and/or surgeons. These functional disorders are presumed to be secondary to altered colonic motor activity, however, the pattern of motor activity of the colorectum in health and in disease has not been fully elucidated to date. This is in part due to the inaccessibility of the organ, and the intermittent and sporadic nature of colonic activity.

The majority of published studies have employed the technique of perfused tube manometry.3–7 The bulky extracorporal apparatus precludes its use for ambulatory studies, requiring the subject to remain sedentary and be placed in the abnormal environment of the laboratory. Stress is known to affect colonic motility8 and laboratory conditions are likely to be more stressful than if a subject is in their home surroundings. The results from these studies therefore may not reflect the normal physiological pattern. Solid-state pressure transducers and portable data recorders have led to the use of ambulatory manometry to study gut motility. There have been relatively few published reports of ambulatory colonic manometry. A two-stage transnasal technique has introduced a catheter to record the motor activity of the whole colon.9 In this pancolonic study only three widely spaced (45 cm) pressure transducers were incorporated into the catheter. These loci were too far apart to allow colonic motor activity to be assessed adequately. Other studies have limited the region of interest to the distal colon, with the catheter being introduced transanally.10,11 Regional functional differences imply differences in motor activity patterns; studies limited to the distal colon cannot therefore be seen as representative of the entire colon. The value of these ambulatory studies is therefore limited. The method of introduction of the manometric probes may also influence the recordings of colonic activity. Bowel preparation affects motor activity,7,12 with a trend for motor activity to be increased in the cleansed bowel, especially with regard to stimulation of high-amplitude propagated contractions (HAPC).7

The aim of this study was to evaluate colorectal motor activity under the most ‘physiological’ conditions practicable. Ambulant studies of unprepared bowel were performed without the use of sedation for catheter placement. A study period of 24 h was used to ensure that phasic and infrequent patterns of motor activity of the colorectum were characterized.

Materials and methods


Ten healthy subjects (6F, 4M; median age 22; range 19–31 years) were studied. Subjects had a normal bowel habit, defined as a stool frequency of ≤3 stools per day and ≥  3 stools per week. The subjects were not on any medication, with the exception that females were able to continue taking the contraceptive pill. Subjects had no gastrointestinal symptoms; there was no history of previous gastrointestinal disorder or gastrointestinal surgery with the exception of appendicectomy for acute appendicitis. Females had a negative pregnancy test. All gave written informed consent. The protocol had ethical approval from the Local Research Ethics Committee of Merton, Sutton and Wandsworth Health Authority.

Measurement of motility

Pressure activity within the colorectum was measured by two custom-made silicone coated catheters (Gaeltec, Isle of Skye, UK). Each pressure transducer catheter contained five solid-state pressure transducers separated at 15-cm intervals. A transnasal catheter (length 3.3 m, outside diameter 3 mm) assessed pressures from the terminal ileum to the splenic flexure, a transanal catheter (length 1.95 m, outside diameter 3 mm) measured pressures in the left colon and rectum. This catheter incorporated a central air channel enabling a latex balloon tied at the distal end of the catheter to facilitate placement. Pressure activity was sampled at 4 Hz and stored on a portable battery powered digital solid-state data logger (MMS Universal Portable System, the Netherlands). The data logger had event marker buttons to allow correlation of activities with recorded information. The data were downloaded onto a personal computer at the termination of the study using a dedicated software program (MMS Universal Portable System) and archived to disk.

Experimental protocol

No bowel preparation was used. After an overnight fast, the transnasal catheter was passed into the stomach. The catheter tip was then passed through the pylorus under fluoroscopic guidance. When the tip had passed the pylorus, 8 mL of air was injected into the balloon to allow bowel motor activity to facilitate movement of the catheter. During this period, the subject was then permitted to eat and drink ad libitum.

Over the next few hours, the subject was instructed to advance the catheter gradually. Fluoroscopy was repeated the following morning to check the position of the catheter tip. The catheter was correctly positioned when the tip had just passed the splenic flexure. If the catheter had not migrated sufficiently aborad the subject was asked to return later for further screening. Once in the correct position the balloon was deflated, and any redundant catheter in the stomach was withdrawn to prevent inadvertent aborad migration during the study period. The catheter was secured firmly to the nose with adhesive tape.

Once the proximal catheter was correctly positioned, the transanal catheter was then introduced into the left colon. No sedation or bowel preparation (e.g. cleansing enema) was used. A thread was tied to the tip of the catheter, and grasped by an Endoclip device (Olympus, Japan)13 that had been passed through the biopsy channel of a colonoscope (Olympus 230 L). The colonoscope was then advanced to the descending colon. The endoclip was attached to a colonic mucosal fold thus securing the catheter. The position of both catheters was then checked with fluoroscopy. The catheters were connected to a portable digital recorder (MMS Universal Portable System), and the study started.

During the study period the subjects were freely ambulant and slept at home. Three standardized meals were provided: a breakfast at 08.00 h provided 400 kcal, and a lunch at 12.30 h and an evening meal at 18.00 h each provided 1000 kcal, with at least 45% of the calories provided as fat. Ingestion of alcohol and caffeine containing drinks was prohibited, other fluids were allowed ad libitum. Smokers were able to continue their normal habits so that they would not experience withdrawal symptoms.

Each subject kept a diary of events. The subject recorded their bedtime, the time of wakening, mealtimes, drinking, passing flatus, urinating, urge to defecate and defecation. The study period was for 24 h, with the aim to record measurements of colonic motility during one period of sleep, the act of defecation and the time after two main meals.

At the end of the study, fluoroscopy was repeated to assess any displacement of the catheters, which were then removed by steady traction.

Data analysis

Qualitative analysis was made by visual inspection of the analogue trace, with special regard to the frequency of HAPC. HAPC were defined when contractile peaks had amplitude of contraction of at least 90 mmHg, were seen to propagate between at least two transducers and had a temporal relationship.

Periods of repeated contractile activity of at least 3 min duration were identified in the colon and rectum. The contractile complexes were then analysed with respect to frequency of contraction, duration of complex, and interval between complexes, with the aid of a computer software program (The Whitechapel Society, London, UK). Regional variations in colonic motor activity were evaluated. Colonic regions were defined as: region 1, caecum, ascending colon and hepatic flexure; region 2, transverse colon and splenic flexure; and region 3, descending and sigmoid colon. Colonic and rectal contractile complexes were also assessed when awake and during the sleeping interval. The relationship of rectal contractile activity to colonic contractile activity and HAPC was assessed. A relationship was deemed to exist if events occurred within 10 min of each other.

The motility index was used as a measure of colonic work, defined as the area under the pressure–time curve per time period. The motility index was calculated for different regions of the colon when awake and during the sleeping interval.

Statistical analysis

Diurnal variation in HAPC frequency was evaluated using the Mann–Whitney U-test. Differences in the mean frequency, duration and interval between episodes of colonic and rectal contractile activity with respect to regional and diurnal variations were compared using the Mann–Whitney U-test. The difference of the motility index when awake and asleep was compared using the Wilcoxon signed rank test.


A total of 232 h of recording was obtained, a mean of 23.2 h per subject. The procedure was well tolerated by the subjects except for a sore throat and nose by the end of the study. Removal of the transnasal catheter caused significant abdominal discomfort in four subjects and nausea in one subject. Distally the transanal catheter was coated with faecal materia; although unpleasant during and immediately after removal, no adverse sequelae were reported. With three of the subjects, the transanal catheter was expelled during defecation. In the remaining seven subjects, removal of the transanal catheter was without problem. Displacement of the catheters occurred to a varying degree: in four of the subjects this resulted in the recording sensors moving beyond the boundaries of the regions of interest. In all studies loss of signal from at least one of the sensors occurred.

High amplitude propagated contractions

Sixty-three HAPC were observed in nine subjects, median (interquartile range) 6/24 h (4.9–8.7 h) (Fig. 1). Aborad propagation was observed with 94% of contractions, while 6% of the HAPC propagated in an orad direction. Diurnal variation was apparent, with higher numbers of HAPC being observed when awake, [6 (5–8)][median (iqr)] than when asleep, [1 (0–1); P = 0.008]. Within 1 hour of waking, 6.3% of HAPC occurred, while 19% occurred within 1 h of a meal.

Figure 1.

A cluster of HAPC in the colon (H) prior to defecation. The HAPC are seen in recording leads 2–6 and occur prior to the sensation of an urge to defecate, followed by defecation. Time bar denotes 8 min.

Low-amplitude colonic motor activity

In total, 453 episodes of colonic contractile activity were identified (Fig. 2). Median frequency of contraction was 4.7 cpm (cycles per minute) (range 4.06–5.46), the median duration of contraction was 6.1 min (4.3–9.5), and median interval between contractions was 32.7 min (17.6–74.1).

Figure 2.

Colonic periodic motor activity in recording leads 1–6 propagating oradly and aboradly in the colon. Time bar denotes 30 min.

Regional variation of colonic motor activity

Median frequency, duration and interval of episodes of colonic contractile activity in the three defined colonic regions are shown in Table 1. Contractile activity from four subjects was excluded (200 episodes of colonic contractile activity) from regional variation analysis due to catheter displacement during the study. A total of 253 episodes of colonic contractile activity were analysed: 120 in region 1;, 72 in region 2; and 61 in region 3. Median frequency of contraction was highest in regions 1 and 2: 5.26 (4.84–5.88) and 5.15 (4.15–6.15) cpm, respectively. No significant differences existed between regions 1 and 2 for duration and interval of episodes of colonic contractile activity. Colonic contractile activity in region 3 had a significantly lower median frequency of contraction [3.94 (3.53–4.36)] cpm than the other colonic regions. The median interval between episodes of colonic contractile activity was shortest in region 2 [29.5 min (15.3–54.7 min)]. Duration of contractions was similar in all three colonic regions.

Table 1.  Frequency, duration and interval of periodic motor activity in the three colonic regions and rectum [Median (iqr)]
RegionFrequency(cpm)Duration (min)Interval (min)
  • *

    Region 1 vs. region 3; P < 0.0001.

  • Region 1 vs. rectum; P < 0.0001.

  • Region 2 vs. region 3; P < 0.0001.

  • §

    Region 2 vs. rectum; P < 0.0001.

  • Region 3 vs. rectum; P < 0.0001.

  • **

    Region 1 vs. rectum; P < 0.0001.

  • ††

    Region 2 vs. rectum; P < 0.0001.

  • ‡‡

    Region 3 vs. rectum; P = 0.0013.

  • §§

    Region 2 vs. region 3; P = 0.03.

15.26 (4.83–5.88)*6.6 (4.8–9.78)**36.2 (18.0–77.3)
25.15 (4.15–6.15)§6.1 (4.4–10.3)††29.5 (15.3–54.7)§§
33.94 (3.53–4.36)*6.15 (4.43–10.4)‡‡42.6 (22.6–99.6)§§
Rectum2.93 (2.42–3.58)§9.45 (5.78–16.7)**††‡‡37.4 (22.1–59.4)

Diurnal variation of colonic motor activity

Colonic contractile activity showed varying degrees of diurnal variation by region. Colonic region 1 did not exhibit diurnal variation of episodes of colonic contractile activity (Table 2). Colonic region 2 demonstrated diurnal variation of the duration of complex. The episodes of colonic contractile activity of colonic region 3 demonstrated diurnal variations in frequency and duration.

Table 2.  Frequency, duration and interval of colorectal periodic motor activity when awake and during the sleeping interval [Median (iqr)]
 Frequency (cpm)Duration (min)Interval (min)
RegionAwakeAsleepP valueAwakeAsleepP valueAwakeAsleepP value
  1. ns, Nonsignificant.

(4.8–5.83) (5.08–6.05) (4.78–9.5) (5.38–10.03) (18.0–72.4) (17.7–81.3) 
(4.15–6.12)(4.17–6.24) (4.6–10.7)(4.03–7.78) (17.0–53.7) (14.0–55.1) 
(3.64–4.43) (3.25–4.02) (4.83–12.5) (3.48–5.65) (21.4–88.1) (33.3–236.4) 
(2.48–3.54) (2.32–3.79) (5.43–18.1) (5.98–14.2) (22.0–60.6)(22.3–57.9) 

Propagation of colonic motor activity

Episodes of colonic contractile activity occurred sporadically, and were seen to propagate in both orad and aborad directions; 39% of episodes of colonic contractile activity propagated.

Rectal motor activity

In total, 171 episodes of rectal contractile activity were identified in nine of the subjects (Fig. 3). Frequency, duration and interval between contractions are shown in Table 1.

Figure 3.

Rectal motor complexes (R) in recording leads 8 and 9 propagating aboradly. Time bar denotes 1 h.

There was no significant diurnal variation in the frequency, duration and interval between contractions of episodes of rectal contractile activity (Table 2).

Median frequency of contraction of episodes of rectal contractile activity was significantly lower than the frequency of contraction of episodes of colonic contractile activity in all of the three colonic regions evaluated. Median duration of episodes of rectal contractile activity was significantly longer than episodes of colonic contractile activity in all of the three colonic regions evaluated.

Relationship of rectal motor activity to colonic events

Thirty-six point six per cent of episodes of rectal contractile activity were preceded by either an episode of colonic contractile activity or a high amplitude propagated contraction.

Diurnal variation in motility index

In all three colonic regions the motility index was significantly higher during the day than at night (P = 0.028). Diurnal variation of colonic and rectal motor activity is shown in Fig. 4.

Figure 4.

Diurnal variation of colorectal motor activity. There is a marked reduction of motor activity at night (N) in comparison to day time (D). Time bar 19.5 h.


This technique allows colorectal motor activity to be evaluated in ambulant subjects who are able to perform many of their usual everyday activities in their normal surroundings. There is no need for sedation or bowel preparation for catheter placement, and the methodology allows simultaneous recordings to be made from the entire colon and rectum. Dinning et al.4and Bampton et al.6 have recently developed a technique for the study of the colorectum using a multilumen perfusion catheter assembly delivered proximally per nasum, which does not require a bowel preparation for placement. Although an advance on earlier perfusion studies, it still has not allowed ambulant studies to be performed. The technique employed in this present study, in which subjects were ambulant, therefore permits phasic colonic motor activity to be studied under the most physiological conditions currently possible.

The technique was not without fault; the subjects were exposed to a low dose of radiation to facilitate catheter placement, catheter displacement was evident, and signal loss from the solid-state transducers was not uncommon. The procedure was well tolerated by subjects; nasal discomfort was the main complaint but did not require termination of the study. Removal of the transnasal catheter was on occasion uncomfortable, and distasteful due to the coating with intestinal contents, but no adverse sequelae were reported. The major advance of this technique was to study subjects under ambulant conditions in their own environment, demonstrating that such studies are not just desirable but are feasible.

The main component of colonic motor activity is low amplitude motor activity. Recent studies suggest an important role for low-amplitude colonic motor activity in the propulsion of luminal contents.5,14 This study demonstrates that the colon exhibits periodic activity. There are sustained periods of rhythmic activity of at least 3 min duration, which resemble phase III of the migrating motor complex (MMC) and the RMC. We suggest that such periods of sustained activity should be termed ‘colonic motor complexes’. Median frequency of contraction was 4.9 cpm (range 2.3–9.5 cpm), which is comparable with the frequency of bursts of contractions identified in previous studies, which showed a frequency of 3–8 cpm.2,7,9 Colonic contractile activity occurred sporadically, and was seen to propagate in both orad and aborad directions. Due to technical difficulties (sensor failure and/or catheter displacement) a quantitative assessment of the propagative attributes of colonic contractile activity was not possible.

Regional variations of colonic contractile activity were apparent. In the more proximal colon comprising region 1 (caecum, ascending colon and hepatic flexure) and region 2 (transverse colon and splenic flexure) there were no significant differences for frequency, duration and interval of colonic contractile activity. The contractile frequency was significantly higher in region 1 and region 2 compared to region 3 (descending and sigmoid colon). This observation is in agreement with those of Kerlin et al. who demonstrated a predominant rhythm of 6 cpm in the ascending colon and 2.5–3.5 cpm in the distal colon.15 The interval between contractions was shortest in the transverse colon and splenic flexure; this was significant compared to region 3, but did not reach significance in comparison to region 1. Feline electrophysiological studies have led to the suggestion that the mid-colon is the site of pacemaking activity,16 and this hypothesis has been supported by the findings in human studies in which the overall dominant electrical frequency components resides in the mid colon.17 The observation of a high frequency of contraction and a short interval between complexes is consistent with the hypothesis that the transverse colon and splenic flexure act as the dominant pacemaker region of the colorectum, however, from the results of this study, a role for pacemaking activity in the proximal colon cannot be discounted. Colonic contractile activity of the transverse and distal colon demonstrated diurnal variation. Significant diurnal variations were present in region 3, with frequency and duration of colonic contractile activity both being reduced during sleep. In region 2, duration of the complex was reduced during the sleeping interval. Colonic region 1 in this study did not exhibit diurnal variation of colonic contractile activity frequency, duration or interval between contractions. The motility index was uniformly reduced across all regions during the sleeping interval; this is at odds with the finding of varying degrees of diurnal variation by region with respect to contractile activity. This may reflect the low number of subjects studied.

The most striking features of colonic motor activity are HAPC, which are thought to be the motor correlate of mass movements. The median number of HAPC observed in this study (6 per 24 h), is similar to that reported previously in both ambulatory and static perfusion studies,1,10 but is less than that reported by Herbst et al., who reported a median frequency of 12 per 24 h18 and Cook et al., who reported a mean of 10.2 per 24 h.3 Herbst concluded that the observed difference between his and earlier studies was methodological, suggesting that the bowel in the subjects he studied still contained faeces (despite the fact that a cleansing enema was used), acting as a stimulus to HAPC development. In this study, no preparation was used at all and the frequency of HAPC observed was similar to that previously reported by Bassotti and Crowell. A wide intersubject variation (1.01–15.3 per 24 h) was observed in normal subjects, as has been previously documented by Cook et al. and Herbst et al., whose ranges were higher at 6–25 per 24 h and 0–40 per 24 h, respectively.3,18 This normal variation in HAPC frequency should be borne in mind when altered HAPC frequency is proposed as a pathophysiological feature of motility disorders, e.g. a decreased HAPC frequency in chronic idiopathic constipation.19 Propagation of HAPC was seen in orad and aborad directions, with the vast majority propagating aboradly. Diurnal variation was apparent, which is in agreement with other studies.1,10 Crowell reported a strong association of HAPC with defecation, with 45% of HAPCs occurring in the hour preceding defecation, 33% occurring within 1 h of waking, and 54% within 2 h after ingestion of a fat meal.10 Such strong associations were not observed in this study; 19% occurred within 1 h of a meal but only 6.3% within 1 h of waking, while 33% of HAPCs preceded defaecation, passing wind or the feeling of an urge to defecate. An explanation for these differences may be purely methodological; in Crowell et al.'s study, solid-state transducers were used to investigate the sigmoid colon, bowel preparation was used and the meal had a higher fat content (60% vs. 45% of calories as fat).10

Ambulant studies have revealed much about the nature of rectal motor activity. The rectum has been shown to exhibit periodic motor activity,20 which has diurnal variation.20 In this study diurnal variations of rectal contractile activity were not seen. Aborad and orad conduction of RMC was seen, confirming another report.11 Contractile frequency in the rectum was significantly lower than the contractile frequency of the colon and the duration of rectal contractile activity was significantly longer than colonic contractile activity in all of the three colonic regions evaluated. Periodic rectal motor activity has been associated temporally with more proximal colonic motor activity, suggesting that periodic rectal motor activity is triggered by the arrival of stool or gas from the colon, and may serve as a braking mechanism to prevent untimely flow of colonic contents.11 A 1 : 1 relationship between rectal contractile activity and either proceeding HAPC or colonic contractile activity was not observed, activities with which one may expect displacement of intracolonic contents. This observation therefore provides evidence against the ‘colonic brake’ hypothesis.


In summary, the methodology employed allowed prolonged study of the entire colorectum without the requirement for any form of bowel preparation. To date this has not been performed in ambulant patients, thus providing unique data for analysis. This advance in methodology allowed colorectal motor activity to be studied under conditions that are the most physiological currently possible. Colonic motor activity exhibits two main components: HAPC and low-amplitude colonic contractile activity. We suggest that periodic low-amplitude colonic contractile activity be termed ‘colonic motor complexes’. HAPC may be responsible for ‘mass movements’, while low-amplitude colonic contractile activity may produce more gradual displacement of intraluminal contents. Periodic motor activity in the colon does not exhibit a 1 : 1 relationship to that seen in the rectum, and the precise relationship of periodic colonic motor activity to that of the small bowel remains to be determined. Intersubject variation as exemplified by HAPC frequency highlights the need for replication and further research to achieve a fuller understanding of normal colorectal motility.