Diurnal Changes in Myoelectric Spiking Activity of the Human Colon

GASTROENTEROLOGY 1985;88:1104-10

Diurnal Changes in Myoelectric Spiking Activity of the Human Colon J.

FREXINOS, 1. BUENO, and

J.

FIORAMONTI

Department of Gastroenterology, Rangueil Hospital and Laboratory of Pharmacology INRA, Toulouse, France

Using an intraluminal probe supporting eight groups of electrodes, the myoelectrical activity of the proximal, transverse, and distal colon was recorded during 24-h sessions in 10 healthy subjects receiving evening and noon meals (>800 kcal) and breakfast «300 kcal), At each colonic site considered, the electromyograms exhibited two kinds of spike bursts: (a) short spike bursts localized at one electrode site and appearing rhythmically at a frequency of 10.6 ± 0.3/min and (b) long spike bursts isolated or propagated orally or aborally, A peculiar pattern consisting of nearly permanent short spike bursts at a rate of 6.7 ± O.4/min was observed at the rectosigmoid junction. Computerized analysis of the duration of each kind of spike burst showed that the long spike burst activity increased by 63%-129% (p < 0.05) during 2 h after each meal (but not after breakfast) at each colonic site. Furthermore, a significant (p < 0.05) decrease in the long spike burst but not short spike burst activity was observed during sleep. These results provide evidence for circadian variations of colonic motility associated with eating and sleeping in the healthy human. Variability is probably the main characteristic of colonic motor function when analyzed by means of transit time and fecal characteristics (1). This fact can be attributed to great fluctuations of the basal motor activity of the colon, already described over periods of recording not exceeding 3 h (2,3). Despite abundant literature, well-defined profiles of colonic motility have not been established. Several factors Received February 27, 1984. Accepted October 18, 1984. Address requests for reprints to: Dr. J. Fioramonti, Laboratoire de Pharmacologie, INRA, 180 chemin de Tournefeuille, 31300 Toulouse, France. This work was supported in part by INSERM grants Nos. 130005, 30, 39 and in part by Fondation pour la Recherche Medicale. The authors thank M. Caussette for constructing the probe, Dr. P. Suduca and Dr. J. Escourrou for placing the probes, and C. Betoulieres for technical assistance. © 1985 by the American Gastroenterological Association 0016-5085/85/$3.30

may be involved: (a) most of the manometric and electromyographic studies only describe the rectosigmoid colon (3,4), not the activity of the whole colon and (b) when the motility of the proximal and distal colon have been recorded simultaneously, only two or three sites have been investigated (5). We have shown that the probability of detecting activity in short segments will be increased using a greater number of recording sites (6). Moreover, 24 h seems to be the minimum recording period in which to establish a colonic motor profile. In previous work performed in dogs (that present a relatively welldefined pattern of colonic contractions) 24 h of recording were necessary to establish a colonic motor profile related to food intake (7). Thus, recording myoelectric spiking activity, which is the electric counterpart of colonic contractions (6), during 24 h at several sites of the colon seems appropriate and necessary to establish a colonic motor profile. The present study was designed to evaluate the myoelectric activity at eight sites along the colon over 24 h using contact electrodes in healthy subjects.

Materials and Methods Subjects The protocol for this study was approved by the Ethical Committee of the Medical Faculty of the University of Toulouse (March 21, 1983). Ten healthy volunteers, aged 25-35 yr, who had one or two bowel movements per day were studied. They were 7 men and 3 women who had not complained of gastrointestinal symptoms, had not taken laxatives for at least 3 yr, and had not taken any other medication for at least 3 wk. The absence of colonic lesions was endoscopically confirmed and the probe was positioned in the colon during colonoscopy.

Abbreviations used in this paper: LSB, long spike burst; SSE, short spike burst.

May 1985

Recording System The probe used to record electric activity was modified from that previously described for recording the activity of the descending and sigmoid colon (8). It was made of a polyvinyl tube, 0.5 cm in diameter and 150 cm in length, and supported eight groups of three electrodes, the groups being 17 cm apart. The first group was located 10 cm from the tip. The three electrodes of each group were placed at 4-mm intervals.' The presence of three electrodes ensured that tht;lte would be a pair suitable for bipolar recording. Each electrode consisted of a ring of nickel-chrome wire (80%-20%, Johnson Matthey Metals Ltd., London, U.K.) fixed around the probe. The Wires leading from the electrodes to the recorder were introduced into the lumen of the tube and exited at the distal end. A thread 180 cm in length was attached to the tip of the tube. Bipolar recordings were made with an eight-channel electroencephalograph (Minihuit Alvar, Paris, France) using a short time constant (0.03 s). The electrical signal arising from four electrode sites was simultaneously picked up on a magnetic tape recorder (Medilog 4-24, Oxford Instrument Company Ltd., U.K.). Recording Session All ~ubjects were fasted for at least 14 h before introduction of the probe. A careful bowel cleansing (tap water enema) was performed 24 and 12 h before placing the probe. The probe was introduced into the colon by means of a colonoscope (CFIBW, Olympus, Corporation of America, New Hyde Park, N.Y.). The thread attached to the tip of the probe was ,introduced into the operator channel of the colonoscope, permitting us to pull the probe to the cecum. The endoscope Was then withdrawn and the position of the probe was noted under fluoroscopy. During colonoscopy, intravenous injections of fentanyl (Janssen Le Brun, Paris, France) (0.1 mg) and diazepam (Roche SA, Neuilly, France) (10 mg) were used as necessary; they were followed by injection of naloxone (Winthrop Lab, Clithy, France) (0.4 mg) at the end. The probe was placed at about 9 AM. To eliminate any possible effects of colohoscopy, air insufflation, enemas, or medications, recording sessions began only at 5 PM, i.e., 7-8 h after placement of the probe. None of the subjects asked to defecate during the recording session. A meal was given at noon, 5 h before the beginning of the recording session. The subjects remained in bed and sleep was not permitted during the day. A 8001000-kcal meal was given at 7 PM the first day. On the second day a continental breakfast «300 kcal) was given at 8 AM and a second meal (800-1000 kcal) at noon. The meals (670 ml, pH 4.8) consisted of vegetables (200 g), potatoes (120 g), meat (100 g). oil (20 g). French cheese (25 g), bread (50 g). one orange or apple, and tap water ad libitum. The meals provided 320-410 kcal of fat, 260-320 kcal of carbohydrate, and 220-270 kcal of protein. The position of the probe was radiographically confirmed 15 h after the beginning of recording. In 2 additional subjects, a simultaneous recording of respiratory movements was performed using a pneumo-

24-HOUR COLONIC SPIKING ACTIVITY

1105

graph belt. placed around the chest. In these subjects we used a probe equipped with a catheter 1.5 mm in diameter exiting at the tip. In order to determine any influ~nce of the presence of fluid in the colon on the character of the electromyogram, recordings were performed during infusion of tap water through the catheter at a rate of 10 mllmin for 15 min.

Analysis of the Signals The signals recorded on the magnetic tape from four electrode sites (ascending, transverse, descending, and sigmoid colon) were quantitatively analyzed using a microcomputerized system previously described in detail (9). Briefly, we distinguished two kinds of colonic spike bursts (6,10) on the electromyogram. The computing system recognized each kind of burst on the basis of its duration and gave the time (minutes and percent of time) occupied by each kind of burst every 60 min for each channel. In order to eliminate a maximum number of artifacts, the signal was filtered before computing (bandpass filter). Only the activity above a variable threshold, depending on the amplitude of spike bursts, was taken into account and continuous activity for more than 45 s was elminiated. Moreover, the number of spike bursts propagated over the whole length of the colon was measuted by visual inspection, of the eight-channel records. Resultswere expressed as mean ± SEM. Spiking activity expressed as a percentage of time at the four sites was compared by variance analysis. The postprandial changes in spiking activity were determined by cOrriparing values obtained for 2 h before and after the meals, and values obtained during the nighttime wete compared with those obtained from 5 to 7 PM (paired t-test).

Results Patterns of Spiking Activity Beoause of an aboral movement of the probe duting withdrawal of the coloIioscope, differences in colon size between the subjects, and displacement of the colon duriIig colonoscopy, the tip of the probe was localized ih the right colon in 4 subjects. and at the hepatic flexure in 6 other subjects; The last group of electrodes was localized in the sigmoid colon ih 7 subjects and in the area of the rectosigmoid junction in 3 other subjects. In all subjects the two kinds of spikes preViously described (6) in the descending and sigmoid colon were observf:ld (Figure 1). The first kind consisted bf short spike bursts (SSBs) lasting 3.1 ± 0.4 sand appearing most often rhythmically at an average rate of 10.6 ± 0.3/min. These bursts always appeared to be localized at one electrode site without any propagation to an adjacent group of electrodes. The second kind of spiking activity consisted of long spike bursts (LS13s) lasting 10.3 ± 3.6 s and could appear in any of three patterns: (a) localized at one electrode

1106

FREXINOS ET AL.

GASTROENTEROLOGY Vol. 88, No.5, Part 1

....

LSB(a) MLSB !501']Y ......................-oJv-......,."...--."v-------r.J'.,...........,...................~rMN~-............__KW---~.~.- ,,-L--'

-'0--]

~---------~----------~~~~--~]

""'4N-NWvI\""""""""'~"'""""""-"-.N\fo.Ny_ _~"" f"'~""""......,...~.......III~..........----,--.-.-'_1"_.'~~] ~~~~v-----_v~--. .n~~~~~~~~~~lr~~----~~~~--] 6

I minute

Figure 1. A representative electromyogram obtained from the right colon (1) to the rectosigmoid junction (8). Short spike bursts (SSBs) are present at a frequency of 11 bursts/min at the site of electrode 2. In the middle of the figure, a long spike burst (LSB) (a) appears in isolation in the right colon; another (h) propagates in oral direction over 34 cm (between sites 6 and 4). On the right side of the tracing a migrating long spike burst (MLSB) propagates from the right colon to the rectosigmoid junction. The latter region shows nearly constant SSB activity, at a frequency of 7 bursts/min.

site and sometimes occurring rhythmically at a rate of about 3/min; (b) propagated over two-five electrode sites, i.e., from 17 to 68 cm in aboral or oral directions, at a speed of 3.9 ± 1.6 cm/s; (c) rapidly propagated in the aboral direction over the whole length of the colon investigated (migrating LSBs at a speed of 9.3 ± 2.4 em/so Simultaneous recording of the respiratory movements in 2 additional subjects did not indicate any correlation between the respiratory movements and the occurrence of spike bursts, which remained unchanged during apnea (Figure 2A). The artifacts induced on the electromyogram when the subject moved in bed appeared simultaneously at all recording sites and were clearly distinguishable from spike bursts (Figure 2A). On the other hand, the presence of a significant volume of fluid in the colon did not alter the characteristics of the electromyographic signals (Figure 2B).

pattern was predominant in that the SSBs occupied 23.8% ± 3.9% of time. In some instances, SSBs appeared for periods of 5-10 min followed by periods of quiescence of 10-20 min. At this level, the

®

l ... j.'u ...... ~

... U·H".'.P" ... ·ltt

"

,

50~v

;"tr--ffl

2~~~~~-~----r-- 1 3

,~l

4~~~~~~~~~,'~~___ ~t~1

®

I minute

I~~~.I~IJ,I-I,

3--------~----------~------------

Comparison of Spiking Activity at Different Colonic Sites Inspection of tracings did not indicate any coordination of spike bursts between the eight sites investigated, except for the migrating LSB which propagated rapidly from the right to the sigmoid colon. The comparison of mean values of spiking activity expressed as LSB or SSB activity indicated no significant difference between the four sites automatically analyzed, except for the 3 subjects with the most distal group of electrodes located in the area of the rectosigmoid junction. At this site, SSBs at a frequency of 6.7 ± 0.4/min were observed. This

4 I minute Figure 2. Electromyograms obtained in 2 additional subjects with simultaneous recording of the respiratory movements over 24 h during colonic infusion of tap water (B). The rate of spike bursts did not correlate with the respiratory rate and remained unchanged during apnea. On the right side of A (arrow) an artifact appearing simultaneously on all the recording channels has been induced by a change in the position of the subject in the bed. The presence of a significant volume of fluid (150 ml infused in 15 min) in the colon (B) did not induce peculiar artifacts on the electromyogram recorded at the end of the infusion.

24-HOUR COLONIC SPIKING ACTIVITY

May 1985

22 20

•

.. 18 E

~

16

•

14

o

o o

•

o Ascending

•

flil!I

I!lIlI Sigmoid

Transverse

1107

•

Oescending

C 12

•u 10 ~

,

Q.

!Ii 8 en

...J

....

6 4

2

o

! I

22 20

.~

'0

18 16

•

14

e•

10

o

E 12 ~ !Ii 8
2

o

5-7

7-9

L--OINNER----J

10-0 0-2 2-4 4-6 " - - - - - NIGHT TIME _ _ _---J

6-8

8-10

L-BREAKFAST

......J

10-12

12-2

L-lUNCH ----J

Figure 3. Spiking activity at four colonic sites during the 2 h before and after the three meals, and during the nighttime. Asterisks indicate a significant difference (p < 0.05) between postprandial and preprandial values in the three meal periods. Nocturnal values are different (p < 0.05) from the values for the first preprandial period, 5-7 PM.

LSB activity only corresponded to migrating LSBs propagated over the whole length of the colon investigated. Consequently, the total duration of LSB activity at this site was significantly (p < 0.05) lower than that observed in upper regions and was not taken into account in the calculations of means of LSB activity at the sigmoid (Figure 3). Effect of Meals on Spiking Activity

In comparison with values from the 2 h preceding the meals, a 63%-129% increase in the total duration of LSB activity was observed at all four colonic regions during the 2 h after lunch and after dinner (Figure 3). This increase started 0-10 min after the beginning of the meal and remained significantly elevated over the following 3 h (Figure 4) .. Moreover, 67% of the migrating LSBs appeared during the 3 hours after each of these two meals. In each

colonic region, the SSB activity was unchanged after the two meals. In the rectosigmoid area, the pattern of nearly permanent SSB activity was also not modified (Table 1). At all sites investigated, breakfast did not induce any change in either LSB or SSB activity (Figures 3 and 4). Spiking Activity During Sleep From 10 PM to 6 AM the lowest values of LSBs (expressed as a percentage of time) were observed (Figure 4). At the four sites analyzed these values were significantly (p < 0.05) lower than those observed during the 2 h (5-7 PM) preceding the dinner (Figure 2). During sleep the migrating LSBs were practically absent (Figure 5). In contrast, no change in SSB activity in any of the regions investigated was observed (Figure 4). The permanent pattern of SSBs in the area of the rectosigmoid junction recorded in 3

1108

GASTROENTEROLOGY Vol. 88, No.5, Part 1

FREXINOS ET AL.

Migrating Long Spike Bursts per hour

20

..

-1!5

4

I

~

f-

3

ai

iii ...J !5

2

DINNER

0

6

+

8

LUNCH

10·

0

2

4

10

6

t

12

2

4

_10

I

o

~

5

f!5

~III iii iii

DINNER

10

0

2

4

6

+

9

l

Dinner

3

5

.,

9

II

t

B reokfast

5

:3 {Lunch

Figure 5. Circadian changes in the number of migrating long spike bursts propagated from the right to sigmoid colon. They appeared mainly during the 4 h after the two meals, and they were almost absent during the night.

0 6 +8

7

10 LUN+CH 2

BREAKFAST

TIME (hours)

Figure 4. Hourly variations of total long spike burst (LSB) and short spike burst (SSfl) activity of the transverse colon in 10 subjects. Compared with values obtained between 5 and 7 PM, the LSB activity (top) significantly increased (asterisks, p < 0.05) during the 3 h after the two meals, and decreased during the nighttime (black line). No significant fluctuation appeared for the SSB activity (bottom).

patients remained unchanged (Table 1). In these 3 patients, however, LSB activity in the more proximal parts of the colon showed the same changes as the 7 other patients after the meals and at nighttime (Table 1).

Discussion The method we used required a careful bowel preparation with several tap water enemas in order to obtain a clean colon for complete colonoscopy. This treatment has been thought to influence the motor activity of the colon. In a recent work (3) no difference in the basal motor act~vity of the distal colon was observed in terms of the mean motility index per 3 h of recording between the cleaned and the unwashed colon. Yet it must be pointed out that the motility index was lower in the unwashed colon in 7 of 10 subjects, suggesting that enemas may illdeed stimulate the basal motor activity (3). The same problems arose' with colonoscopy, air insuffla-

Table 1. Changes in Spiking Activity (Percent of Time) in the Three Subjects With the Last Group of Electrodes Located in the Rectosigmoid Area Before meal Descending colon LSBs SSBs Rectosigmoid C SSBs

After meal

Nighttime

17-19 h

10-12 h

19-21 h

12-14 h

22-0 h

0-2 h

2-4 h

4-6 h

8.6 ± 1,7 2.3 ± 2.2

7,2 ± 1.2 3.1 ± 3.2

15,5 ± 2.10 2.4 ± 2.7

16,1 ± 2.6 0 4.2 ± 3.5

3.7 ± 1.2 b 2.6 ± 2.4

3.2 ± 0.6" 3.3 ± 3.7

4.3 ± 0.9 b 2.7 ± 2.7

3.5 ± 1.l b 4.4 ± 3.6

23.0 ± 3.6

23.4 ± 2.8

26.1 ± 4.7

24.5 ± 4.1

22.6 ± 4.5

22.2 ± 4.2

25.2 ± 4.4

26.6 ±

~.8

LSBs, long spike bursts; SSBs, short spike bursts. Mean ± SEM. 0 Significantly different (p < 0.05) from vqlues during 2 h before the meal. b Significantly different (p < 0.05) from values obtained at 17-19 h. C The LSBs, which only corresponded in this area to migrating long spike bursts, were not taken into account.

May 1985

tion, and drug injections. These constraints must be noted although (a) the time between the last enema (12 h) or the colonoscopy (8 h) and the beginning of the recording session seems to be sufficient to eliminate the majority of these influencing factors and (b) water infusion did not alter the character of the record. It is obvious, however, that a truly physiologic basal condition was not obtained. All these methodo logic and technical questions must be taken into account and caution should be exercised in the interpretation of the results. The levels of LSB and SSB activity obtained in this study, however, are similar to those previously obtained in the descending colon without enemas and air insufflation (6), suggesting that the degree of fullness of the colon has a limited influence on these values. The two kinds of spiking activity described here corresponded to those previously reported by Bueno et al. (6) and by other investigators (4,11). The SSBs correspond to the discrete electrical response activity reported by Sarna et al. (11) and the rhythmic spike potentials reported by Schang and Devroede (4). The LSBs correspond to the continuous electrical response activity and the contractile electrical complexes of Sarna et al. (11), to the sporadic spike potentials of Schang and Devroede (4), and probably to the migrating spike bursts described by Christensen et al. (12,13) in the cat colon in vitro. Previous experiments (6) showed that LSBs corresponded to phasic contractions of large amplitude and SSBs were associated with contractions of small amplitude. Our study indicates the presence of a peculiar pattern at the rectosigmoid junction consisting of predominant SSB activity. This pattern corresponds to that described by Wright et al. (14) in the same area in healthy subjects, consisting of slow waves at a frequency of 6/min superimposed with spike bursts similar to the SSBs. This fact indicates that the rectosigmoid does not reflect the activity of the whole colon in both basal and postprandial states. The almost continuous SSBs and corresponding mechanical activity (6) indicates increased activity of the rectosigmoid and confirms the presence of a hyperactive segment that may act as a "brake" at this level (15). This study provides evidence for circadian variations in colonic myoelectric activity associated with eating and sleeping. The postprandial increase in LSB activity confirms previous findings (4) obtained in the sigmoid colon with postprandial recording sessions that did not exceed 2 h after the meal. Our long-duration recording session, however, allowed us to record a response that lasted about 3 h. This. response may be correlated with the postprandial increase in the flow of the liquid phase of digesta

24-HOUR COLONIC SPIKING ACTIVITY

1109

through the ileum, which begins 30 min after the meal and lasts ~3 h with a maximum at 1.5 h (16). The early (0-30 min) increase in spiking activity after the beginning of the meal can be attributed to the gastrocolic response observed during the first 20 min after a meal in the distal colon (17). The colonic myoelectric response, however, does not correspond to the arrival of the solid phase into the colon, as the peak of breath Hz appears only 3 h after a meal and lasts more than 5 h (16). On the other hand, because lunch and dinner (but not breakfast) elicited an increase in LSB activity, our results confirm that a minimum caloric load >300 kcal is needed to induce a colonic motor response (18). Our results indicate no postprandial changes in the pattern of SSB activity in the rectosigmoid area. This fact is not in accordance with the colonic response to food intake previously observed for a short period after a meal (60-90 min) in the area of the rectosigmoid junction (14,17,18). This discrepancy with previous studies may be attributed to our analysis: we analyzed 1-h intervals, which are probably inadequate to detect a short-duration (20 min) increase in spiking activity, and we analyzed the percentage of time occupied by spiking activity, not the number of spike potentials (14,17,18). Contradictory data are available on colonic motility during sleep. It has been suggested that sleep depresses colon motor activity (19); some experimental data confirm this opinion (20), but others do not provide evidence for change in colonic motility (21) or note a great variability of motility in the rectosigmoid during sleep (5). Our quantitative analysis indicates a significant decrease in LSB activity during the night.

References 1. Wyman JV, Heaton KW, Wicks ACB. Variability of colonic function in healthy subjects. Gut 1978;19:146-50. 2. Chaudhary NA, Truelove Sc. Human colonic motility. A comparative study of normal subjects, patients with ulcerative colitis and patients with irritable colon syndrome. 1. Resting patterns of motility. Gastroenterology 1961;40:1-46. 3. Dinoso VP, Murthy SNS, Goldstein J, Rosner B. Basal motor activity of the distal colon: a reappraisal. Gastroenterology 1983;85:637-42. 4. Schang JC, Devroede G. Fasting and postprandial myoelectric spiking activity in the human sigmoid colon. Gastroenterology 1983;85:637-42. 5. Kerlin P, Zinsmeister A, Philips S. Motor responses to food of the ileum, proximal colon, and distal colon of healthy humans. Gastroenterology 1983;84:762-70. 6. Bueno L, Fioramonti J, Ruckebusch y, Frexinos J, Coulom P. Evaluation of colonic myoelectric activity in health and functional disorders. Gut 1980;21:480-5. 7. Fioramonti J, Bueno 1. Diurnal changes in colonic motor profile in conscious dogs. Dig Dis Sci 1983;28:257-64. 8. Fioramonti J, Bueno L, Frexinos J. Sonde endoluminale pour

1110

9.

10.

11.

12.

13. 14.

15.

FREXINOS ET AL.

l'exploration electromyographique de la motricite colique chez l'homme. Gastroenterol Clin BioI 1980;4:546-50. Latour A, Bueno L, Fioramonti J. Quantitative measurement of human colonic electrical activity by a micro-computerized system. Int J Biomed Com put 1983;14:7-16. Fioramonti J, Bueno L. Motor activity in the large intestine of the pig related to dietary fibre and retention time. Br J Nutr 1980;43:155-62. Sarna SK, Waterfall WE, Bardakjian BL, Lind JF. Types of human colonic electrical activities recorded postoperatively. Gastroenterology 1981;81:61-70. Christensen J, Anuras S, Hauser RL. Migrating spike bursts and electrical slow waves in the cat colon: effect of sectioning. Gastroenterology 1974;66:240-7. Christensen J. Myoelectric control of the colon. Gastroenterology 1975;68:601-9. Wright SH, Snape WJ, Battle W, Cohen S, London RL. Effect of dietary components on gastrocolonic response. Am J Physiol 1980;238:G228-32. Chowdhury AR, Dinoso VP, Lorber SH. Characterization of a

GASTROENTEROLOGY Vol. 88, No.5, Part 1

16.

17.

18.

19. 20.

21.

hyperactive segment in the rectosigmoid junction. Gastroenterology 1976;71:548-8. Kerlin P, Philips S. Differential transit of liquids and solid residue through the human ileum. Am J Physiol 1983;245: G38-43. Sun EA, Snape WJ, Cohen S, Renny A. The role of opiate receptors and cholinergic neurons in the gastrocolonic response. Gastroenterology 1982;82:689-93. Snape WJ, Matarazzo SA, Cohen S. Effect of eating and gastrointestinal hormones on human colonic myoelectric and motor activity. Gastroenterology 1978;75:373-8. Hagihara PF, Griffen WOo Physiology of the colon and rectum. Surg Clin North Am 1972;52:797-805. Adler HF, Atkinson AJ, Ivy AC. A study of the motility of the human colon. An explanation of dysynergia of the colon, or of the "unstable colon." Am J Dig Dis 1941;8:197-202. Posey EL, Bargen JA. Observations of normal and abnormal human intestinal motor function. Am J Med Sci 1951;221:1020.