The American Journal of Clinical Nutrition 34: OCTOBER
© 1981 American Society for Clinical Nutrition
1981, pp. 2111-2116. Printed in U.S.A.
Limits of predicting gastrointestinal transit time
from other measures of bowel function1’2
J. L. Slavin, M.S., C. T Sempos, M.S., P. M. Brauer, M.S., and J. A. Marlett,3 Ph.D.
ABSTRACT The two aims of these experiments were first, to examine the relationships
between mean daily stool weight, mean weight per stool, frequency of defecation, dye transit time,
dietary fiber intake, and transit time of small radiopaque pellets, and second, to determine the
ability of these measures of bowel function, singly or in combination, to predict gastrointestinal
transit time ofthe pellets. Variables were observed simultaneously in 13 healthy women consuming
controlled low and high cellulose diets. All except one of the correlation coefficients between the
measures of bowel function were significant. Stepwise regression analysis retained log stool weight,
fiber and dye transit to predict log of the mean transit time of all pellets (R2 = 71.3); the same three
variables were selected to predict log of the transit time of 80% of the pellets (R2 = 59.4). Mean
daily stool weight explained about 50% of the variation for both measures of transit, while dye
transit and dietary fiber explained the rest. Log transformation of either stool weight or pellet
transit time or both variables improved the prediction by about 10%. These results suggest that
prediction of pellet gastrointestinal transit time from other measures of bowel function may be
limited and is influenced significantly by the fiber level of the diet. Am. J. Clin. Nuir. 34: 2111-
2116, 1981.
KEY WORDS Gastrointestinal transit time, stool weight, dietary fiber, correlation of bowel
functions
Introduction
Gastrointestinal transit time is mainly a
colonic event ( 1) and has been suggested as
a measure of overall colonic function (2). The
availability ofa method to estimate gut transit
that could be conveniently and easily used in
a field study of a large number of individuals
would be useful in the study of the role of
dietary fiber in bowel function. Correlations
between transit time and other measures of
bowel function, that are simpler to determine,
suggest it may be possible to predict transit
time from one or more of these measures.
The transit time of small radiopaque pellets
has been correlated with mean daily stool
weight (2-7), dye (8), and frequency of defe-
cation, although the correlation with defeca-
tion frequency was significant in only one (9)
of three (9-11) studies. However, rate of pas-
sage of dye has been criticized because it
cannot be conveniently quantitated in feces
(12) and may not always be visible (13).
Further, dietary fiber appears to influence
the correlations of pellet transit time with
stool weight (7), defecation frequency (9), or
dye (8).
Stool collection and access to x-ray facili-
ties are necessary to conveniently quantitate
the pellets, whereas mean daily stool weight
would require only stool collection. Deter-
mination of defecation frequency and transit
time of dye would not require stool collection
from competent, adequately instructed
adults. The aims of this study, therefore, were
to examine the correlations between these
various measures of bowel function assessed
in subjects consuming constant diets and to
determine whether stool weight, defecation
frequency, and dye transit time, singly or in
From the Department of Nutritional Sciences, Uni-
versity of Wisconsin-Madison, 1300 Linden Drive, Mad-
ison, WI 53706.
2 Supported in part by University of Wisconsin Amer-
ican Cancer Society Institutional Grant IN-35Q-6, Na-
tional Institutes of Health Grant AM2l7l2, and The
College of Agricultural and Life Sciences, University of
Wisconsin-Madison.
Author to whom requests for reprints should be
addressed.
2111
by guest on April 16, 2012 www.ajcn.org Downloaded from