- Email: [email protected]
Simplified assessment of segmental colonic transit
GASTROENTEROLOGY
1987;92:4o-7
Simplified Assessment of Segmental Colonic Transit AMANDA ALAN
M. R.
METCALF,
ZINSMEISTER,
SIDNEY
F.
PHILLIPS,
ROBERT
L.
MAcCARTY,
ROBERT W. BEART, and BRUCE G. WOLFF Gastroenterology Unit, Department of Surgery, Section of Medical Research Department of Radiology, Mayo Medical School, Rochester, Minnesota
Transit times of radiopaque markers through the human gut were measured by published techniques and compared with a simplified method. Three sets of distinctive markers were ingested by 24 healthy persons on 3 successive days. In the first part of the study, daily abdominal x-rays were taken and individual stools were collected for radiography. Mouthto-anus transits were assessed from the fecal output of markers and mean colonic and segmental colonic transits were calculated from the daily radiographs. These established methods were then compared with estimates of total colonic and segmental transits based on a single abdominal film, taken on the fourth day. The single-film technique correlated well with values obtained from the previous, but more inconvenient, methods. Using the simpler approach, colonic transit was assessed in 49 additional healthy subjects, for a total group of 73. Total colonic transit was 35.0 2 2.1 h [mean -+ SEJ; segmental transits were I 1.3 t 1.1 h for the right colon, 11.4 k 1.4 hfor the left colon, and 12.4 t 1.1 h for the rectosigmoid. Men had significantly shorter transits for the whole colon than did women (p < O.O5), and this difference was apparent to some extent in the right (p = 0.06) and left colon (p = 0.07) but not in the rectosigmoid. Age did not influence transit significantly nor did a small dose of supplemental fiber. The technique is simple, convenient for clinical usage, and reduces the exposure to radiation to acceptable levels. There should be a role for Received February 28, 1986. Accepted June 12, 1986. Address requests for reprints to: Sidney F. Phillips, M.D., Gastroenterology Unit, Mayo Clinic, 200 Southwest First Street, Rochester, Minnesota 55905. This work was supported in part by grants AM32121, 34988, and RR585 from the National Institutes of Health, Bethesda, Maryland. The authors thank the staff of the Mayo GCRC (B. Koenig) and the Gastroenterology Unit (D. Warren) for their expert help. 0 1987 by the American Gastroenterological Association 0016-5085/87/$3.50
Statistics,
and
this approach in the evaluation selected patients.
of colonic transit in
Symptoms attributed to disorders of colonic motility are commonly encountered complaints in clinical practice and, as such, colonic dysmotility is considered an important health problem (1,~). Despite the general acceptance of this premise, clinical disorders of colonic motility remain obscure; indeed, most approaches that describe and quantitate colonic motility experimentally are not readily applicable to the clinical context. Thus, of the various methods of measuring colonic transit time, this being one index of motility, only those tests that involve the oral administration of a “marker” and its subsequent progress through the gut have been feasible clinically (3-7). However, even these have potential drawbacks. Those approaches that observe the appearance of markers in stools are rather inconvenient and, in addition, only measure transit times through the entire gut, from mouth to anus (3,8). They give no indication of segmental transit times within the large intestine. Yet, certain disorders of gastrointestinal motility may involve localized segments of the large bowel (5,6); indeed, techniques that quantify segmental colonic transit have been described (7). These have not been used widely, because of the inconvenience of daily x-rays of the abdomen and the exposure to radiation thus entailed. The aim of this study was twofold: first, to validate a simple and practical measurement of colonic transit utilizing radiopaque markers and a single abdominal film and second, to evaluate the effects of age, gender, and modest variations in dietary fiber on total and segmental colonic transit. Abbreviations used in this paper: MCT, mean colonic MMAT, mean mouth-to-anus transit.
transit;
January
TECHNIQUE FOR SEGMENTAL COLONIC TRANSIT
1987
Materials and Methods ProtocoI The protocol was approved by the Institutional Review Board and Radiation Safety Committee of Mayo Clinic, and all subjects gave written informed consent. Criteria for participation included a usual stool frequency of between 3 stools daily and 3 stools weekly, no history of gastrointestinal disease, and no use of medications known to affect gastrointestinal motility. All volunteers were asked to maintain their usual diet for the duration of the study, and were randomly assigned to ingest no additional fiber or to receive a daily supplement of fiber (10 g of Fiberall, Rydelle Laboratories, Inc., Racine, Wis.) to test for any effects of minor variations in dietary fiber. Subjects were asked to avoid the use of laxatives and all women capable of childbearing had a negative serum human choriogonadotrophin test to exclude the possibility of pregnancy. Three types of radiopaque markers (types l-3) were prepared; 2 x 6-mm circular (type 1) and 6 X 6-mm semicylindrical markers (type 2) were cut from a 16F radiopaque Levine tube (Tomac 19947-9616, KIOR, specific gravity 1.46); 1 x 6-mm cylindrical (type 3) markers were cut from radiopaque pediatric tubing (OD 1 mm, specific gravity 1.33). The size of each marker was such that 20 of one type could be enclosed within two 00 gelatin capsules. Subjects ingested 20 markers at 9:00 AM on each of 3 consecutive days; the three types of markers were
41
always ingested in the same sequence. Abdominal x-rays were obtained using a high-kilovoltage fast-film technique to reduce the amount of radiation exposure (estimated surface exposure = 0.08 mrad per film). The different marker types were easily distinguishable on both abdominal films (Figure 1) and radiographs of the stools. Localization of markers on abdominal films relied on the identification of bony landmarks and gaseous outlines as described by Arhan and colleagues (7). In the absence of clear outlines of the bowel, markers located to the right of the vertebral spinous processes above a line from the fifth lumbar vertebrae to the pelvic outlet were assigned to the right colon. Markers to the left of the vertebral spinous processes and above an imaginary line from the fifth lumbar vertebrae to the anterior superior iliac crest were assigned to the left colon. Markers inferior to a line from the pelvic brim on the right and the superior iliac crest on the left were judged to be in the rectosigmoid and rectum. However, if bowel outlines clearly showed a pelvic cecum, a transverse colon, or a large sigmoid loop above the fifth lumbar vertebrae, markers were judged to be in the anatomic segment based on the gaseous outlines. For each film, markers ingested on days 1, 2, and 3 were assigned to one of the colonic segments. In a separate analysis, for comparison purposes only, the colon was divided into six different segments (see below). Validation of Technique: Three Markers
Comparison
of
In the initial phase, volunteers (10 men and 14 women) without a past history of gastrointestinal disease. and with a normal YLVV~ frequency, were recruited in two age ranges: 53.5 yr of age (n = 15) or 260 yr of age (n = 9). All subjects were randomized to ingest no supplemental fiber or to receive supplemental fiber (10 g of Fiberall) for the duration of the study. The three sets of markers were ingested on successive days, and 24 h after ingestion of the first set of markers an abdominal x-ray was obtained. This was repeated at 24-h intervals until all markers had been passed. Stools were also collected from day 1 until all markers had been passed. Each specimen was kept in an individual plastic container and labeled with the time and date of defecation. Radiographs of the stools were used to count the number and types of markers passed. Influence
of Age, Gender,
and
Fiber
In this phase of the study, 49 additional healthy volunteers (25 women and 24 men), 20 yr of age and older (range 21-69 yr; female median 28 yr, male median 31 yr) were studied. Subjects ingested 20 of each kind of marker at 24-h intervals, and a single abdominal film was obtained 24 h after ingestion of the last type of marker (day 4). All subjects were randomized to ingest no supplemental fiber or to receive supplemental fiber (10 g of Fiberall) for the duration of the study. Calculations Figure
1. Plain radiograph types of marker.
of the abdomen
showing
the three
Mouth-to-Anus Transit Time. In the initial group of 24 subjects, data from x-rays of individual stools were
42 METCALF ET AL.
GASTROENTEROLOGY Vol. 92, No. 1
used to calculate the mean mouth-to-anus transit (MMAT) times for each marker, using the formula described by Hinton et al. (3). MMAT = i i=1
xiti I
2 xi, i=1
(11
where ti = time from ingestion of markers, xi = number of markers (of type 1,2,or 3) passed at time ti, and n is the total number of stools. The average MMAT time, from the individual transit times for the three types of marker in each person, was calculated using the following formula: Average MMAT = $MMAT, + MMAT, + MMAT,),
(2)
where MMATl, MMATz, MMATB are the MMAT times for the type 1, type 2, and type 3 markers, respectively. Mean colonic and segmental colonic transit time (using all daily films), Daily abdominal films were used to calculate the mean total colonic transit time and the mean segmental colonic transit times for each type of marker using the method of Arhan et al. (7).
where MCT = mean colonic transit, N = total number of a particular type of marker given. In Equation (3), ni = number of each type of marker present on a film taken at time ti,~(t(i+l) - t(i_1)) = time interval between successive films, and j is the number of abdominal films taken. As the abdominal films were always taken at 24-h intervals, and the total number of each type of marker was kept constant at 20, the above formula can be simplified to MCT = _!_ i
20 f=l
ni x 24 = 1.2 i
i=1
Average MCT = ;(MCT, + MCTz + MCTB),
(41
where MCT1, MCT2, and MCTB are the MCT times calculated for the type 1, type 2, and type 3 markers, respectively. Mean Colonic and Segmental Transit (single-film estimate). Assuming that the colon handles the different types of markers similarly and yet independently, the number of markers present on a single x-ray taken on the fourth study day after ingestion of 20 of three kinds of markers at 24-h intervals should be equivalent to the total number of markers present on the first three sequential abdominal films after a single bolus of markers: 3
is1
Methods/Data
Analysis
Estimated overall mouth-to-anus transits for each of three types of marker were obtained for subjects with complete stool collections usiug Equation (1). The differences between these three estimates (type 1 - type 2, and type 2 - type 3) were analyzed using two-way analysis of covariance (age as the covariate) to evaluate possible systematic differences in marker movement due to sex or fiber, or both. In addition, for each subject, the type of marker yielding the maximal and minimal overall gut transit times was noted. Then the proportion of subjects with type 1, 2, or 3 as the maximal, and as the minimal, was calculated to check for any consistent serial effects that may have resulted from the order of marker ingestion. The three estimates of overall colonic transit and those of segmental transit [Equation (3)] using multiple abdominal films were analyzed similarly. Associations between MMAT and MCT for different marker types were summarized and compared. In addition, differences between single-film estimates [Equation (5)] and the average multiple-film estimate [Equation (4)] were evaluated for possible systematic effects of gender and fiber using two-way analysis of covariance (with age as the covariate). Actual mean overall colonic and segmental estimated transit times based on multiple films (n = 24) were evaluated for differences between men and women and for the effects of fiber supplementation, after adjusting for possible age effects using two-way analysis of covariance. The age-adjusted mean transit times for each group were then computed. Finally, the distributions of single-film estimates of overall colonic and segmental transit time were graphed using 73 subjects. The corresponding 95th percentile (and 95% confidence interval for this percentile) were estimated for overall and segmental transit.
ni.
The average of the MCT times obtained for each marker was calculated for each individual subject using Equation (4) and is hereafter referred to as the average MCT time.
MCT = 1.2 z
Statistical
ni = 1.2(nl + nZ + n3) = 1.2(N1 + Nz + Nd, (51
where N1 = No. of the first type of marker present, N2 = No. of the second type of marker present, N3 = No. of the third type of marker present.
Results Validation Markers
of Technique:
Comparison
of
Mouth-to-anus transit. Twenty-one of 24 subjects provided complete stool collections, thus allowing MMAT times to be calculated using Equation (1). The MMAT times calculated for the different markers correlated well with each other. For all subjects, the mean difference in MMAT times calculated for type 1 and type 2 markers was -0.3 ? 2.2 h; they correlated well (r = 0.89). The mean difference in MMAT times calculated for type 2 markers and type 3 markers was + 2.7 +- 3.0 h, and the correlation was r = 0.69. A test for trends (linear and quadratic) in transit times as a function of day of ingestion failed to indicate any overall significance (i.e., trends were no different from zero over all subjects using a onesample t-test). Thus, the variability in transit times between the different markers in an individual presumably reflected true day-to-day variation in
TECHNIQUE
January 1987
FOR SEGMENTAL
COLONIC TRANSIT
43
60
60
60
Average mean
Average of mean colonic transit
colonic transit
2
4
6
8
10
12
40
2:: 0
14
2. Mean colonic transit plotted against stool frequency in 21 healthy subjects; transit was not related to bowel habit.
40
2
60
8
0
4th day estimate
Weekly stool frequency Figure
20
Figure
3. Comparison of average mean colonic transit times calculated for each marker type from daily abdominal x-rays (ordinate) and estimated transit, based on a single abdominal film (day 4). The line is drawn for x = Y,
mouth-to-anus transit in that individual. The median coefficient of variation within subjects (i.e., variation among three marker type estimates as a percentage of the overall mean mouth-to-anus estimate] was 17% (inner quartile range of 9%-25%). For the entire group, the average value of MMAT -I SE (all markers) was 53.3 + 3.7 h. Analysis of the pairwise differences between estimates of overall mouth-to-anus transit based on the three different marker types failed to detect any effects of age, gender, and fiber. The overall mean differences were not statistically different from zero (one-sample t-test). Colonic transit. Based on multiple films, MCT times for the whole colon were well correlated for each type of marker. Overall, average differences between MCT calculated for type 1 and type 2 markers were 0.9 t 2.1 h, and between type 2 and type 3 markers, 2.1 ? 2.6 h. Analysis of the pairwise differences in estimated total colonic transit from the three marker types failed to reveal any trends with respect to the day of ingestion. The median withinTable 1. Mean Colonic
Transit
subject coefficient of variation for multiple-film estimates (variation among marker types as a percentage of overall mean x-ray estimate) was 20% (inner quartile range of 12%35%). The average MCT correlated well (r = 0.89) with the average MMAT although, as expected, the average mouth-to-anus transit was consistently longer than the average MCT (colon only). The average MCT for the entire group was 39.9 h or 13.1 h shorter than the average mouth-to-anus transit time. Stool frequency during the study period did not reflect differences in mouth-to-anus transit time (not shown) or MCT time (Figure 2). Although older subjects had longer transit times, the difference was not statistically significant. For purposes of statistical comparison the age-adjusted mean values for men vs. women and fiber vs. no fiber are given in Table 1. The average MCT tended to be shorter in men than women (p < 0.05). The segmental transit times demonstrated these differences to be
Times Mean colonic transit times (h)” Right colon
Variable
Left colon
Rectosigmoid
Total colon
Gender Men (n = 10) Women (n = 14) Fiber
Supplemented Men (n = 6) Women (n = 9)
9.2 * 2.4 19.0 2 2.0
9.9 k 2.3 15.3 r 2.0
14.3 t 2.0 12.5 '- 1.7
33.3 k 4.3 46.9 2 3.7
9.6 -c 3.0 14.9 f 2.4
9.0 5 2.9 12.6 It 2.3
12.6 +- 2.6 12.4 + 2.1
31.3 2 5.5 39.9 f 4.4
8.7 t 3.6 23.2 -c 3.3
10.7 k 3.5 18.0 k 3.1
16.0 + 3.1 12.6 k 2.6
35.4 k 6.6 53.8 * 5.9
Not supplemented Men (n = 4) Women (n = 5) a Average values for 24 healthy
persons.
Values are from multiple
daily abdominal
films; age-adjusted.
44
METCALF ET AL.
GASTROENTEROLOGY Vol. 92, No.
Right colon Figure
4. Comparisons of average mean segmental colonic transits calculated for each marker type from daily abdominal x-rays (ordinates) and estimated transits, based on single abdominal film on fourth day (abscissas). The lines drawn are for x=y.
(h) -0
40
20
-0
40
20
4th day estimate
most apparent in the right colon (Table 1); the left colon showed lesser differences, and sigmoid colon transit times were similar. The average MCT was shorter in the group receiving supplemental fiber, although these differences were not statistically significant. Estimated colonic transit time, based on calculations from the single fourth-day film, correlated well (r = 0.87) with the average MCT [Figure 3). The
0
3
6
9
12
15
16
21 24
27
30
Single film right colonic transit
33
36
39
estimate,
42
45
46
0
40
(h)
6
9
12
15
16
21 24
27
30
33
36
39
42
45
46
h
20 fD
-
z
15
-
z ,” u 0 IO .k
-
‘fZ J? P
if IO
20
Single tilm left colonic transit estimate,
s
;$ m z E
3
h
rr
8 0” E 15
-0
mean difference in colonic transit times so calculated was -0.6 + 1.7 h, the mean estimated (fourthday film) colonic transit being 39.3 h. The estimated mean segmental colonic transits calculated from the fourth-day film alone, correlated well with the average mean segmental colonic transit calculated from all films (Figure 4). The mean difference in transit for each segment between those calculated from daily films and from the fourth-day estimate was small for
25
20
Rectosigmoid
Left colon
-
95%
5-
5-
n0
Figure
3
39
42
Single film sigmoid colonic transit estimate,
6
9
12
15
16
21 24
27
30
33
36
h
45
46
0
5
IO
15
20
25
30
35
40
45
50
55
60
Single film colonic transit estimate,
65
70
75
80
h
5. Histograms of segmental and total transit time estimates in 73 subjects based on single fourth-day abdominal film. Median values were (a) right colon 10.8 h, (b) left colon 8.4 h, (c) sigmoid colon 9.6 h, (d) total colon 32.4 h. Values for the 96th percentile (with confidence limits) are shown.
TECHNIQUE FOR SEGMENTAL COLONIC TRANSIT
January 1987
45
Table 2. Effects of Variables on Colonic Transit in 73 Healthy Person? Colonic transit (h) Left colon
Right colon
Variable Gender Men (n = 34) Women (n = 39)
Total colon
Rectosigmoid
8.9 + 1.1 13.3 ir 1.6
8.7 t 1.5 13.7 + 2.1
13.0 '- 1.7 11.8 + 1.6
30.7 " 3.0 38.8 -r-2.9
Fiber Supplemented (n = 39) Not supplemented (n = 34)
10.9 ? 1.6 11.7 5 1.1
11.0 k 1.6 11.9 k 2.2
12.9 2 1.7 11.7 + 1.5
34.7 t 2.9 35.3 ? 3.1
Age >40 yr (n = 21) <40 yr (n = 52)
9.4 + 1.6 12.0 k 1.3
11.8 + 2.8 11.2 2 1.5
11.8 " 1.6 12.6 k 1.4
33.0 k 3.9 35.8 2 2.5
a Estimated
from a single, fourth-day
film.
each segment. These mean differences were -2.5 2 1.1 h in the right colon, +2.5 + 1.5 h in the left colon, and -0.7 ? 1.1 h in the rectosigmoid.
significant effect of fiber or of age on transit could be detected.
times
Discussion Phase II As only fourth-day films were obtained in additional subjects, these results were combined with estimates from fourth-day films in the first phase, for a total of 73 subjects. The MCT time had an average of 35.0 k 2.1 h (Figure 5); the median value was 32.4 h and the 95th percentile value was 68.0 h. The averages of segmental transit times were as follows: right colon 11.3 + 1.1 h, left colon 11.4 + 1.4 h, and rectosigmoid 12.4 & 1.1 h. Figure 5 also illustrates the distribution of transit times for each segment. Men had shorter (p < 0.05) transit times for the entire colon compared to women. This difference was apparent in the right (p = 0.06) and left colon (p = O.O7), but not in the rectosigmoid (Table 2). No Men
Women
-
Mean colonic transit 30.7 k 3.0 h
=
Mean colonic transit 38.8 + 2.9 h
=
Figure 6. Proportion of total colonic transit time spent in cecum and ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, sigmoid colon, and rectum in 73 normal subjects.
Our major aim was to establish a simple technique, using radiopaque markers, for quantifying segmental colonic transit. In particular, we hoped by reducing the number of abdominal films required, to validate a method which would be convenient for the clinical setting. The multiplemarker, single-film technique provided a good estimate of total and segmental colonic transit in normal subjects; moreover, it was convenient and reduced the radiation exposure to acceptable levels. For the entire study, a single high-kilovoltage, fast-film exposed subjects to approximately one-sixteenth the radiation of the previous technique which uses multiple standard abdominal films; the high-kilovoltage approach delivered approximately one-fourth the radiation of the usual abdominal x-ray. Thus, the method has acceptable precision, is simple, and minimizes radiation exposure; these features make application of this test to clinical studies a realistic possibility. Unlike previously published simplified techniques which have focused on whole gut or total colon transit (8,9), the present method measures both total and segmental colonic transit. The inherent variability in colonic transit documented in the initial phase of this study suggests that in a clinical setting only major differences from normality can be accepted as a significant finding. Although individuals with unusually prolonged coionic transit will be identified by a single film, an additional abdominal x-ray will be necessary if precise values demonstrating a clear abnormality in transit are desired. The multiple-marker, single-film technique only monitors the progress of markers in the colon for 72 h. Although this provides a good
46
METCALF ET AL.
estimate for persons with MCTs of <72 h, it is clear that the approach will underestimate transit in individuals with MCTs longer than 72 h. Thus, retention of all 60 markers on day 4 yields an estimated transit [from Equation (5)] of 72 h, a value at the upper limit of normal (Figure 5). However, as all markers are still in the colon, actual transit time must be longer than 72 h. To quantify the degree of abnormality more precisely, or to localize an abnormality to one coionic segment, a further film (on day 7) would then be necessary. The concept of extending the period of observation is even more important in individuals with an abnormality confined to one colonic segment, such as rectal inertia or “outlet” obstruction. The inherent variability in day-to-day transit within colonic segments may falsely prolong transit in one segment despite a normal overall transit time. For example, if transit in the right colon on the third day of study was faster than on previous days, markers may accumulate disproportionately in the distal colon. Indeed, Figure 5 shows that some persons had left colonic transits of 36 h or longer, despite a normal overall colonic transit. Verification of prolonged segmental transit would require confirmation by a film on day 7. If transit were indeed prolonged, many markers would still be in the distal segments, but if transit were within normal limits few or no markers would remain. Transit through the colon as measured by this technique was not related to age or to a modest variation in the intake of dietary fiber. Others have reported wide intrinsic variations in intestinal transit (10)and a lack of response to fiber using similar amounts of supplement (ll,l2).However, studies using larger doses of fiber have demonstrated decreased whole gut transit (13,14), though the effect was not seen in all subjects. Whether large doses of supplemental fiber would decrease colonic transit time is not answered by this study, but small variations in dietary fiber clearly do not affect colonic transit time. Gender appeared to influence transit, with men having shorter transit times than women, though overlap between the groups was considerable. Mean orocecal transit times could be estimated from MMAT and MCT. The average value of 13 h was relatively long. Although marker ingestion was not standardized in relation to meals, ingestion at 9 AM was presumably in proximity to the morning meal; others have demonstrated a marked delay in the gastric emptying of undigestible solids taken in association with a meal (15). While variation in orocecal transit might influence the assumptions underlying the formulas used to calculate colonic transit, any such effects would be small in compar-
GASTROENTEROLOGY Vol. 92, No. 1
ison to the overall transit. In each subject in this study, all markers were clearly in the colon 24 h after ingestion. Any major functional differences between the segments of the colon were not reflected by our measurements of transit. Thus, transit times through the three segments we chose to analyze were similar. These were the divisions of the large bowel used by others in comparable studies (7), but separation of the colon into these segments is arbitrary. An alternate breakdown of the results was also made (Figure 6) by dividing the colon into cecum and ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, and rectosigmoid segment. We analyzed the proportion of total colonic transit time spent in each of these segments. Transit appeared largely proportional to the anticipated length and volume of the segment, except for prolonged storage in the rectosigmoid. Clinical experience with disorders of colonic transit suggests that some abnormalities may be segmental and, as such, may benefit from a specific therapeutic approach (5). Our technique, using multiple markers and a single abdominal film, provides a good estimate of segmental colonic transit times. It is simple, convenient, and minimizes radiation exposure. This test should be most applicable to persons with normal or prolonged transit. Use in patients with rapid transit, although possible, would require significant modifications of the times at which markers are administered and abdominal films are obtained. Note added in manuscript. Since the acceptance of this manuscript, a report was published by Chaussade et al. (16) describing results in 22 volunteers using the same technique as described here. The mean transit times (+2 SD] for right, left, and rectosigmoid colons were 6.9 + 15.6,9.1 t 20.6,and 18.4 + 25.0 h, respectively.
References 1. Almy TP. Digestive disease as a national
2.
3.
4.
5.
problem. II. A white paper by the Gastroenterological Association. Gastroenterology 1969;53:821-33. Harvey RF, Salih SY, Read AE. Organic and functional disorders in 2000 gastroenterology outpatients. Lancet 1983$:632-l. Hinton J, Lennard-Jones J, Young A. A new method of studying gut transit times using radio-opaque markers. Gut 1969;10:842-7. Eastwood HD. Bowel transit studies in the elderly: radioopaque markers in the investigation of constipation. Gerontol Clin 1972;14:154-9. Mantelli H, Devroede G, Arhan P, Duguay E. Mechanisms of idiopathic constipation: outlet obstruction. Gastroenterology 1978;75:623-31.
January 1987
6. Waite A, Devroede G, Duranceau A, et al. Constipation with colonic inertia: a manifestation of systemic disease? Dig Dis Sci 1983;28:1025-33. 7. Arhan P, Devroede G, Jehannu B, et al. Segmental colonic transit time. Dis Colon Rectum 1981;24:625-9. 8. Cummings JH, Wiggins HS. Transit through the gut measured by analysis of a single stool. Gut 1976;17:219-23. 9. Becker U, Elsborg L. A new method for the determination of gastrointestinal transit times. Stand J Gastroenterol 1979; 14:355-g. 10. Wyman JB, Heaton KW, Manning AP, Wicks ACB. Variability of colonic function in healthy subjects. Gut 1978;19:146-50. 11. Durrington PN, Manning AP, Bolton CH, Hartog M. Effect of pectin on serum lipids and lipoproteins, whole-gut transittime and stool weight. Lancet 1976;i:394-6. 12. Baird IM, Walters RL, Davies PS, Hill MJ, Drasar BS, Southgate DAT. The effect of two dietary fiber supplements on
TECHNIQUE FOR SEGMENTAL COLONIC TRANSIT
47
gastrointestinal transit, stool weight and frequency, and bacterial flora, and fecal bile acids in normal subjects. Metabolism 1977;26:117-28. 13. Cummings TH, Jenkins DJA, Wiggins HS. Measurement of the mean transit time of dietary residue through the human gut. Gut 1976;17:210-8. 14. Wyman JB, Heaton KW, Manning AP, Wicks ACB. The effect on intestinal transit and the feces of raw and cooked bran in different doses. Am J Clin Nutr 1976;29:1474-9. 15. Mojaverian P, Ferguson RK, Vlasses PH, et al. Estimation of gastric residence time of the Heidelberg capsule in humans: effect of varying food composition. Gastroenterology 1985; 89:392-7. 16. Chaussade S, Roche H, Khyara A, Couturier D, Guerre J. Mesure du temps de transit colique (TTC): description et validation dune nouvelle technique. Gastroenterol Clin Biol 1986;10:385-9.
1987;92:4o-7
Simplified Assessment of Segmental Colonic Transit AMANDA ALAN
M. R.
METCALF,
ZINSMEISTER,
SIDNEY
F.
PHILLIPS,
ROBERT
L.
MAcCARTY,
ROBERT W. BEART, and BRUCE G. WOLFF Gastroenterology Unit, Department of Surgery, Section of Medical Research Department of Radiology, Mayo Medical School, Rochester, Minnesota
Transit times of radiopaque markers through the human gut were measured by published techniques and compared with a simplified method. Three sets of distinctive markers were ingested by 24 healthy persons on 3 successive days. In the first part of the study, daily abdominal x-rays were taken and individual stools were collected for radiography. Mouthto-anus transits were assessed from the fecal output of markers and mean colonic and segmental colonic transits were calculated from the daily radiographs. These established methods were then compared with estimates of total colonic and segmental transits based on a single abdominal film, taken on the fourth day. The single-film technique correlated well with values obtained from the previous, but more inconvenient, methods. Using the simpler approach, colonic transit was assessed in 49 additional healthy subjects, for a total group of 73. Total colonic transit was 35.0 2 2.1 h [mean -+ SEJ; segmental transits were I 1.3 t 1.1 h for the right colon, 11.4 k 1.4 hfor the left colon, and 12.4 t 1.1 h for the rectosigmoid. Men had significantly shorter transits for the whole colon than did women (p < O.O5), and this difference was apparent to some extent in the right (p = 0.06) and left colon (p = 0.07) but not in the rectosigmoid. Age did not influence transit significantly nor did a small dose of supplemental fiber. The technique is simple, convenient for clinical usage, and reduces the exposure to radiation to acceptable levels. There should be a role for Received February 28, 1986. Accepted June 12, 1986. Address requests for reprints to: Sidney F. Phillips, M.D., Gastroenterology Unit, Mayo Clinic, 200 Southwest First Street, Rochester, Minnesota 55905. This work was supported in part by grants AM32121, 34988, and RR585 from the National Institutes of Health, Bethesda, Maryland. The authors thank the staff of the Mayo GCRC (B. Koenig) and the Gastroenterology Unit (D. Warren) for their expert help. 0 1987 by the American Gastroenterological Association 0016-5085/87/$3.50
Statistics,
and
this approach in the evaluation selected patients.
of colonic transit in
Symptoms attributed to disorders of colonic motility are commonly encountered complaints in clinical practice and, as such, colonic dysmotility is considered an important health problem (1,~). Despite the general acceptance of this premise, clinical disorders of colonic motility remain obscure; indeed, most approaches that describe and quantitate colonic motility experimentally are not readily applicable to the clinical context. Thus, of the various methods of measuring colonic transit time, this being one index of motility, only those tests that involve the oral administration of a “marker” and its subsequent progress through the gut have been feasible clinically (3-7). However, even these have potential drawbacks. Those approaches that observe the appearance of markers in stools are rather inconvenient and, in addition, only measure transit times through the entire gut, from mouth to anus (3,8). They give no indication of segmental transit times within the large intestine. Yet, certain disorders of gastrointestinal motility may involve localized segments of the large bowel (5,6); indeed, techniques that quantify segmental colonic transit have been described (7). These have not been used widely, because of the inconvenience of daily x-rays of the abdomen and the exposure to radiation thus entailed. The aim of this study was twofold: first, to validate a simple and practical measurement of colonic transit utilizing radiopaque markers and a single abdominal film and second, to evaluate the effects of age, gender, and modest variations in dietary fiber on total and segmental colonic transit. Abbreviations used in this paper: MCT, mean colonic MMAT, mean mouth-to-anus transit.
transit;
January
TECHNIQUE FOR SEGMENTAL COLONIC TRANSIT
1987
Materials and Methods ProtocoI The protocol was approved by the Institutional Review Board and Radiation Safety Committee of Mayo Clinic, and all subjects gave written informed consent. Criteria for participation included a usual stool frequency of between 3 stools daily and 3 stools weekly, no history of gastrointestinal disease, and no use of medications known to affect gastrointestinal motility. All volunteers were asked to maintain their usual diet for the duration of the study, and were randomly assigned to ingest no additional fiber or to receive a daily supplement of fiber (10 g of Fiberall, Rydelle Laboratories, Inc., Racine, Wis.) to test for any effects of minor variations in dietary fiber. Subjects were asked to avoid the use of laxatives and all women capable of childbearing had a negative serum human choriogonadotrophin test to exclude the possibility of pregnancy. Three types of radiopaque markers (types l-3) were prepared; 2 x 6-mm circular (type 1) and 6 X 6-mm semicylindrical markers (type 2) were cut from a 16F radiopaque Levine tube (Tomac 19947-9616, KIOR, specific gravity 1.46); 1 x 6-mm cylindrical (type 3) markers were cut from radiopaque pediatric tubing (OD 1 mm, specific gravity 1.33). The size of each marker was such that 20 of one type could be enclosed within two 00 gelatin capsules. Subjects ingested 20 markers at 9:00 AM on each of 3 consecutive days; the three types of markers were
41
always ingested in the same sequence. Abdominal x-rays were obtained using a high-kilovoltage fast-film technique to reduce the amount of radiation exposure (estimated surface exposure = 0.08 mrad per film). The different marker types were easily distinguishable on both abdominal films (Figure 1) and radiographs of the stools. Localization of markers on abdominal films relied on the identification of bony landmarks and gaseous outlines as described by Arhan and colleagues (7). In the absence of clear outlines of the bowel, markers located to the right of the vertebral spinous processes above a line from the fifth lumbar vertebrae to the pelvic outlet were assigned to the right colon. Markers to the left of the vertebral spinous processes and above an imaginary line from the fifth lumbar vertebrae to the anterior superior iliac crest were assigned to the left colon. Markers inferior to a line from the pelvic brim on the right and the superior iliac crest on the left were judged to be in the rectosigmoid and rectum. However, if bowel outlines clearly showed a pelvic cecum, a transverse colon, or a large sigmoid loop above the fifth lumbar vertebrae, markers were judged to be in the anatomic segment based on the gaseous outlines. For each film, markers ingested on days 1, 2, and 3 were assigned to one of the colonic segments. In a separate analysis, for comparison purposes only, the colon was divided into six different segments (see below). Validation of Technique: Three Markers
Comparison
of
In the initial phase, volunteers (10 men and 14 women) without a past history of gastrointestinal disease. and with a normal YLVV~ frequency, were recruited in two age ranges: 53.5 yr of age (n = 15) or 260 yr of age (n = 9). All subjects were randomized to ingest no supplemental fiber or to receive supplemental fiber (10 g of Fiberall) for the duration of the study. The three sets of markers were ingested on successive days, and 24 h after ingestion of the first set of markers an abdominal x-ray was obtained. This was repeated at 24-h intervals until all markers had been passed. Stools were also collected from day 1 until all markers had been passed. Each specimen was kept in an individual plastic container and labeled with the time and date of defecation. Radiographs of the stools were used to count the number and types of markers passed. Influence
of Age, Gender,
and
Fiber
In this phase of the study, 49 additional healthy volunteers (25 women and 24 men), 20 yr of age and older (range 21-69 yr; female median 28 yr, male median 31 yr) were studied. Subjects ingested 20 of each kind of marker at 24-h intervals, and a single abdominal film was obtained 24 h after ingestion of the last type of marker (day 4). All subjects were randomized to ingest no supplemental fiber or to receive supplemental fiber (10 g of Fiberall) for the duration of the study. Calculations Figure
1. Plain radiograph types of marker.
of the abdomen
showing
the three
Mouth-to-Anus Transit Time. In the initial group of 24 subjects, data from x-rays of individual stools were
42 METCALF ET AL.
GASTROENTEROLOGY Vol. 92, No. 1
used to calculate the mean mouth-to-anus transit (MMAT) times for each marker, using the formula described by Hinton et al. (3). MMAT = i i=1
xiti I
2 xi, i=1
(11
where ti = time from ingestion of markers, xi = number of markers (of type 1,2,or 3) passed at time ti, and n is the total number of stools. The average MMAT time, from the individual transit times for the three types of marker in each person, was calculated using the following formula: Average MMAT = $MMAT, + MMAT, + MMAT,),
(2)
where MMATl, MMATz, MMATB are the MMAT times for the type 1, type 2, and type 3 markers, respectively. Mean colonic and segmental colonic transit time (using all daily films), Daily abdominal films were used to calculate the mean total colonic transit time and the mean segmental colonic transit times for each type of marker using the method of Arhan et al. (7).
where MCT = mean colonic transit, N = total number of a particular type of marker given. In Equation (3), ni = number of each type of marker present on a film taken at time ti,~(t(i+l) - t(i_1)) = time interval between successive films, and j is the number of abdominal films taken. As the abdominal films were always taken at 24-h intervals, and the total number of each type of marker was kept constant at 20, the above formula can be simplified to MCT = _!_ i
20 f=l
ni x 24 = 1.2 i
i=1
Average MCT = ;(MCT, + MCTz + MCTB),
(41
where MCT1, MCT2, and MCTB are the MCT times calculated for the type 1, type 2, and type 3 markers, respectively. Mean Colonic and Segmental Transit (single-film estimate). Assuming that the colon handles the different types of markers similarly and yet independently, the number of markers present on a single x-ray taken on the fourth study day after ingestion of 20 of three kinds of markers at 24-h intervals should be equivalent to the total number of markers present on the first three sequential abdominal films after a single bolus of markers: 3
is1
Methods/Data
Analysis
Estimated overall mouth-to-anus transits for each of three types of marker were obtained for subjects with complete stool collections usiug Equation (1). The differences between these three estimates (type 1 - type 2, and type 2 - type 3) were analyzed using two-way analysis of covariance (age as the covariate) to evaluate possible systematic differences in marker movement due to sex or fiber, or both. In addition, for each subject, the type of marker yielding the maximal and minimal overall gut transit times was noted. Then the proportion of subjects with type 1, 2, or 3 as the maximal, and as the minimal, was calculated to check for any consistent serial effects that may have resulted from the order of marker ingestion. The three estimates of overall colonic transit and those of segmental transit [Equation (3)] using multiple abdominal films were analyzed similarly. Associations between MMAT and MCT for different marker types were summarized and compared. In addition, differences between single-film estimates [Equation (5)] and the average multiple-film estimate [Equation (4)] were evaluated for possible systematic effects of gender and fiber using two-way analysis of covariance (with age as the covariate). Actual mean overall colonic and segmental estimated transit times based on multiple films (n = 24) were evaluated for differences between men and women and for the effects of fiber supplementation, after adjusting for possible age effects using two-way analysis of covariance. The age-adjusted mean transit times for each group were then computed. Finally, the distributions of single-film estimates of overall colonic and segmental transit time were graphed using 73 subjects. The corresponding 95th percentile (and 95% confidence interval for this percentile) were estimated for overall and segmental transit.
ni.
The average of the MCT times obtained for each marker was calculated for each individual subject using Equation (4) and is hereafter referred to as the average MCT time.
MCT = 1.2 z
Statistical
ni = 1.2(nl + nZ + n3) = 1.2(N1 + Nz + Nd, (51
where N1 = No. of the first type of marker present, N2 = No. of the second type of marker present, N3 = No. of the third type of marker present.
Results Validation Markers
of Technique:
Comparison
of
Mouth-to-anus transit. Twenty-one of 24 subjects provided complete stool collections, thus allowing MMAT times to be calculated using Equation (1). The MMAT times calculated for the different markers correlated well with each other. For all subjects, the mean difference in MMAT times calculated for type 1 and type 2 markers was -0.3 ? 2.2 h; they correlated well (r = 0.89). The mean difference in MMAT times calculated for type 2 markers and type 3 markers was + 2.7 +- 3.0 h, and the correlation was r = 0.69. A test for trends (linear and quadratic) in transit times as a function of day of ingestion failed to indicate any overall significance (i.e., trends were no different from zero over all subjects using a onesample t-test). Thus, the variability in transit times between the different markers in an individual presumably reflected true day-to-day variation in
TECHNIQUE
January 1987
FOR SEGMENTAL
COLONIC TRANSIT
43
60
60
60
Average mean
Average of mean colonic transit
colonic transit
2
4
6
8
10
12
40
2:: 0
14
2. Mean colonic transit plotted against stool frequency in 21 healthy subjects; transit was not related to bowel habit.
40
2
60
8
0
4th day estimate
Weekly stool frequency Figure
20
Figure
3. Comparison of average mean colonic transit times calculated for each marker type from daily abdominal x-rays (ordinate) and estimated transit, based on a single abdominal film (day 4). The line is drawn for x = Y,
mouth-to-anus transit in that individual. The median coefficient of variation within subjects (i.e., variation among three marker type estimates as a percentage of the overall mean mouth-to-anus estimate] was 17% (inner quartile range of 9%-25%). For the entire group, the average value of MMAT -I SE (all markers) was 53.3 + 3.7 h. Analysis of the pairwise differences between estimates of overall mouth-to-anus transit based on the three different marker types failed to detect any effects of age, gender, and fiber. The overall mean differences were not statistically different from zero (one-sample t-test). Colonic transit. Based on multiple films, MCT times for the whole colon were well correlated for each type of marker. Overall, average differences between MCT calculated for type 1 and type 2 markers were 0.9 t 2.1 h, and between type 2 and type 3 markers, 2.1 ? 2.6 h. Analysis of the pairwise differences in estimated total colonic transit from the three marker types failed to reveal any trends with respect to the day of ingestion. The median withinTable 1. Mean Colonic
Transit
subject coefficient of variation for multiple-film estimates (variation among marker types as a percentage of overall mean x-ray estimate) was 20% (inner quartile range of 12%35%). The average MCT correlated well (r = 0.89) with the average MMAT although, as expected, the average mouth-to-anus transit was consistently longer than the average MCT (colon only). The average MCT for the entire group was 39.9 h or 13.1 h shorter than the average mouth-to-anus transit time. Stool frequency during the study period did not reflect differences in mouth-to-anus transit time (not shown) or MCT time (Figure 2). Although older subjects had longer transit times, the difference was not statistically significant. For purposes of statistical comparison the age-adjusted mean values for men vs. women and fiber vs. no fiber are given in Table 1. The average MCT tended to be shorter in men than women (p < 0.05). The segmental transit times demonstrated these differences to be
Times Mean colonic transit times (h)” Right colon
Variable
Left colon
Rectosigmoid
Total colon
Gender Men (n = 10) Women (n = 14) Fiber
Supplemented Men (n = 6) Women (n = 9)
9.2 * 2.4 19.0 2 2.0
9.9 k 2.3 15.3 r 2.0
14.3 t 2.0 12.5 '- 1.7
33.3 k 4.3 46.9 2 3.7
9.6 -c 3.0 14.9 f 2.4
9.0 5 2.9 12.6 It 2.3
12.6 +- 2.6 12.4 + 2.1
31.3 2 5.5 39.9 f 4.4
8.7 t 3.6 23.2 -c 3.3
10.7 k 3.5 18.0 k 3.1
16.0 + 3.1 12.6 k 2.6
35.4 k 6.6 53.8 * 5.9
Not supplemented Men (n = 4) Women (n = 5) a Average values for 24 healthy
persons.
Values are from multiple
daily abdominal
films; age-adjusted.
44
METCALF ET AL.
GASTROENTEROLOGY Vol. 92, No.
Right colon Figure
4. Comparisons of average mean segmental colonic transits calculated for each marker type from daily abdominal x-rays (ordinates) and estimated transits, based on single abdominal film on fourth day (abscissas). The lines drawn are for x=y.
(h) -0
40
20
-0
40
20
4th day estimate
most apparent in the right colon (Table 1); the left colon showed lesser differences, and sigmoid colon transit times were similar. The average MCT was shorter in the group receiving supplemental fiber, although these differences were not statistically significant. Estimated colonic transit time, based on calculations from the single fourth-day film, correlated well (r = 0.87) with the average MCT [Figure 3). The
0
3
6
9
12
15
16
21 24
27
30
Single film right colonic transit
33
36
39
estimate,
42
45
46
0
40
(h)
6
9
12
15
16
21 24
27
30
33
36
39
42
45
46
h
20 fD
-
z
15
-
z ,” u 0 IO .k
-
‘fZ J? P
if IO
20
Single tilm left colonic transit estimate,
s
;$ m z E
3
h
rr
8 0” E 15
-0
mean difference in colonic transit times so calculated was -0.6 + 1.7 h, the mean estimated (fourthday film) colonic transit being 39.3 h. The estimated mean segmental colonic transits calculated from the fourth-day film alone, correlated well with the average mean segmental colonic transit calculated from all films (Figure 4). The mean difference in transit for each segment between those calculated from daily films and from the fourth-day estimate was small for
25
20
Rectosigmoid
Left colon
-
95%
5-
5-
n0
Figure
3
39
42
Single film sigmoid colonic transit estimate,
6
9
12
15
16
21 24
27
30
33
36
h
45
46
0
5
IO
15
20
25
30
35
40
45
50
55
60
Single film colonic transit estimate,
65
70
75
80
h
5. Histograms of segmental and total transit time estimates in 73 subjects based on single fourth-day abdominal film. Median values were (a) right colon 10.8 h, (b) left colon 8.4 h, (c) sigmoid colon 9.6 h, (d) total colon 32.4 h. Values for the 96th percentile (with confidence limits) are shown.
TECHNIQUE FOR SEGMENTAL COLONIC TRANSIT
January 1987
45
Table 2. Effects of Variables on Colonic Transit in 73 Healthy Person? Colonic transit (h) Left colon
Right colon
Variable Gender Men (n = 34) Women (n = 39)
Total colon
Rectosigmoid
8.9 + 1.1 13.3 ir 1.6
8.7 t 1.5 13.7 + 2.1
13.0 '- 1.7 11.8 + 1.6
30.7 " 3.0 38.8 -r-2.9
Fiber Supplemented (n = 39) Not supplemented (n = 34)
10.9 ? 1.6 11.7 5 1.1
11.0 k 1.6 11.9 k 2.2
12.9 2 1.7 11.7 + 1.5
34.7 t 2.9 35.3 ? 3.1
Age >40 yr (n = 21) <40 yr (n = 52)
9.4 + 1.6 12.0 k 1.3
11.8 + 2.8 11.2 2 1.5
11.8 " 1.6 12.6 k 1.4
33.0 k 3.9 35.8 2 2.5
a Estimated
from a single, fourth-day
film.
each segment. These mean differences were -2.5 2 1.1 h in the right colon, +2.5 + 1.5 h in the left colon, and -0.7 ? 1.1 h in the rectosigmoid.
significant effect of fiber or of age on transit could be detected.
times
Discussion Phase II As only fourth-day films were obtained in additional subjects, these results were combined with estimates from fourth-day films in the first phase, for a total of 73 subjects. The MCT time had an average of 35.0 k 2.1 h (Figure 5); the median value was 32.4 h and the 95th percentile value was 68.0 h. The averages of segmental transit times were as follows: right colon 11.3 + 1.1 h, left colon 11.4 + 1.4 h, and rectosigmoid 12.4 & 1.1 h. Figure 5 also illustrates the distribution of transit times for each segment. Men had shorter (p < 0.05) transit times for the entire colon compared to women. This difference was apparent in the right (p = 0.06) and left colon (p = O.O7), but not in the rectosigmoid (Table 2). No Men
Women
-
Mean colonic transit 30.7 k 3.0 h
=
Mean colonic transit 38.8 + 2.9 h
=
Figure 6. Proportion of total colonic transit time spent in cecum and ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, sigmoid colon, and rectum in 73 normal subjects.
Our major aim was to establish a simple technique, using radiopaque markers, for quantifying segmental colonic transit. In particular, we hoped by reducing the number of abdominal films required, to validate a method which would be convenient for the clinical setting. The multiplemarker, single-film technique provided a good estimate of total and segmental colonic transit in normal subjects; moreover, it was convenient and reduced the radiation exposure to acceptable levels. For the entire study, a single high-kilovoltage, fast-film exposed subjects to approximately one-sixteenth the radiation of the previous technique which uses multiple standard abdominal films; the high-kilovoltage approach delivered approximately one-fourth the radiation of the usual abdominal x-ray. Thus, the method has acceptable precision, is simple, and minimizes radiation exposure; these features make application of this test to clinical studies a realistic possibility. Unlike previously published simplified techniques which have focused on whole gut or total colon transit (8,9), the present method measures both total and segmental colonic transit. The inherent variability in colonic transit documented in the initial phase of this study suggests that in a clinical setting only major differences from normality can be accepted as a significant finding. Although individuals with unusually prolonged coionic transit will be identified by a single film, an additional abdominal x-ray will be necessary if precise values demonstrating a clear abnormality in transit are desired. The multiple-marker, single-film technique only monitors the progress of markers in the colon for 72 h. Although this provides a good
46
METCALF ET AL.
estimate for persons with MCTs of <72 h, it is clear that the approach will underestimate transit in individuals with MCTs longer than 72 h. Thus, retention of all 60 markers on day 4 yields an estimated transit [from Equation (5)] of 72 h, a value at the upper limit of normal (Figure 5). However, as all markers are still in the colon, actual transit time must be longer than 72 h. To quantify the degree of abnormality more precisely, or to localize an abnormality to one coionic segment, a further film (on day 7) would then be necessary. The concept of extending the period of observation is even more important in individuals with an abnormality confined to one colonic segment, such as rectal inertia or “outlet” obstruction. The inherent variability in day-to-day transit within colonic segments may falsely prolong transit in one segment despite a normal overall transit time. For example, if transit in the right colon on the third day of study was faster than on previous days, markers may accumulate disproportionately in the distal colon. Indeed, Figure 5 shows that some persons had left colonic transits of 36 h or longer, despite a normal overall colonic transit. Verification of prolonged segmental transit would require confirmation by a film on day 7. If transit were indeed prolonged, many markers would still be in the distal segments, but if transit were within normal limits few or no markers would remain. Transit through the colon as measured by this technique was not related to age or to a modest variation in the intake of dietary fiber. Others have reported wide intrinsic variations in intestinal transit (10)and a lack of response to fiber using similar amounts of supplement (ll,l2).However, studies using larger doses of fiber have demonstrated decreased whole gut transit (13,14), though the effect was not seen in all subjects. Whether large doses of supplemental fiber would decrease colonic transit time is not answered by this study, but small variations in dietary fiber clearly do not affect colonic transit time. Gender appeared to influence transit, with men having shorter transit times than women, though overlap between the groups was considerable. Mean orocecal transit times could be estimated from MMAT and MCT. The average value of 13 h was relatively long. Although marker ingestion was not standardized in relation to meals, ingestion at 9 AM was presumably in proximity to the morning meal; others have demonstrated a marked delay in the gastric emptying of undigestible solids taken in association with a meal (15). While variation in orocecal transit might influence the assumptions underlying the formulas used to calculate colonic transit, any such effects would be small in compar-
GASTROENTEROLOGY Vol. 92, No. 1
ison to the overall transit. In each subject in this study, all markers were clearly in the colon 24 h after ingestion. Any major functional differences between the segments of the colon were not reflected by our measurements of transit. Thus, transit times through the three segments we chose to analyze were similar. These were the divisions of the large bowel used by others in comparable studies (7), but separation of the colon into these segments is arbitrary. An alternate breakdown of the results was also made (Figure 6) by dividing the colon into cecum and ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, and rectosigmoid segment. We analyzed the proportion of total colonic transit time spent in each of these segments. Transit appeared largely proportional to the anticipated length and volume of the segment, except for prolonged storage in the rectosigmoid. Clinical experience with disorders of colonic transit suggests that some abnormalities may be segmental and, as such, may benefit from a specific therapeutic approach (5). Our technique, using multiple markers and a single abdominal film, provides a good estimate of segmental colonic transit times. It is simple, convenient, and minimizes radiation exposure. This test should be most applicable to persons with normal or prolonged transit. Use in patients with rapid transit, although possible, would require significant modifications of the times at which markers are administered and abdominal films are obtained. Note added in manuscript. Since the acceptance of this manuscript, a report was published by Chaussade et al. (16) describing results in 22 volunteers using the same technique as described here. The mean transit times (+2 SD] for right, left, and rectosigmoid colons were 6.9 + 15.6,9.1 t 20.6,and 18.4 + 25.0 h, respectively.
References 1. Almy TP. Digestive disease as a national
2.
3.
4.
5.
problem. II. A white paper by the Gastroenterological Association. Gastroenterology 1969;53:821-33. Harvey RF, Salih SY, Read AE. Organic and functional disorders in 2000 gastroenterology outpatients. Lancet 1983$:632-l. Hinton J, Lennard-Jones J, Young A. A new method of studying gut transit times using radio-opaque markers. Gut 1969;10:842-7. Eastwood HD. Bowel transit studies in the elderly: radioopaque markers in the investigation of constipation. Gerontol Clin 1972;14:154-9. Mantelli H, Devroede G, Arhan P, Duguay E. Mechanisms of idiopathic constipation: outlet obstruction. Gastroenterology 1978;75:623-31.
January 1987
6. Waite A, Devroede G, Duranceau A, et al. Constipation with colonic inertia: a manifestation of systemic disease? Dig Dis Sci 1983;28:1025-33. 7. Arhan P, Devroede G, Jehannu B, et al. Segmental colonic transit time. Dis Colon Rectum 1981;24:625-9. 8. Cummings JH, Wiggins HS. Transit through the gut measured by analysis of a single stool. Gut 1976;17:219-23. 9. Becker U, Elsborg L. A new method for the determination of gastrointestinal transit times. Stand J Gastroenterol 1979; 14:355-g. 10. Wyman JB, Heaton KW, Manning AP, Wicks ACB. Variability of colonic function in healthy subjects. Gut 1978;19:146-50. 11. Durrington PN, Manning AP, Bolton CH, Hartog M. Effect of pectin on serum lipids and lipoproteins, whole-gut transittime and stool weight. Lancet 1976;i:394-6. 12. Baird IM, Walters RL, Davies PS, Hill MJ, Drasar BS, Southgate DAT. The effect of two dietary fiber supplements on
TECHNIQUE FOR SEGMENTAL COLONIC TRANSIT
47
gastrointestinal transit, stool weight and frequency, and bacterial flora, and fecal bile acids in normal subjects. Metabolism 1977;26:117-28. 13. Cummings TH, Jenkins DJA, Wiggins HS. Measurement of the mean transit time of dietary residue through the human gut. Gut 1976;17:210-8. 14. Wyman JB, Heaton KW, Manning AP, Wicks ACB. The effect on intestinal transit and the feces of raw and cooked bran in different doses. Am J Clin Nutr 1976;29:1474-9. 15. Mojaverian P, Ferguson RK, Vlasses PH, et al. Estimation of gastric residence time of the Heidelberg capsule in humans: effect of varying food composition. Gastroenterology 1985; 89:392-7. 16. Chaussade S, Roche H, Khyara A, Couturier D, Guerre J. Mesure du temps de transit colique (TTC): description et validation dune nouvelle technique. Gastroenterol Clin Biol 1986;10:385-9.