Content uploaded by Stuart Andrew Shaw Craig
Author content
All content in this area was uploaded by Stuart Andrew Shaw Craig on Apr 24, 2018
Content may be subject to copyright.
ABSTRACT
Background: Previous studies showed that polydextrose has
physiologic effects similar to those of dietary fiber.
Objective: Ingestion of 4, 8, and 12 g polydextrose/d was stud-
ied to determine the physiologic effects in Chinese subjects.
Design: In a placebo-controlled, randomized, double-blind
study, we evaluated the effects of polydextrose ingestion on clin-
ical biochemistry indexes, glycated hemoglobin, glucose toler-
ance, the glycemic index, bowel function, stool weight and pH,
short-chain fatty acid production, fecal microflora, and cecal
mucosa cell proliferation.
Results: Polydextrose had no significant effect on blood bio-
chemistry indexes. Ingestion of 12 g polydextrose plus 50 g
glucose resulted in a glycemic index of 89% (compared with a
glycemic index of 100% after ingestion of 50 g glucose).
Bowel function (frequency and ease of defecation) improved
significantly and there were no reports of abdominal distention,
abdominal cramps, diarrhea, or hypoglycemia. Fecal weight
(wet and dry) increased and fecal pH decreased proportionally
to polydextrose intake. Short-chain fatty acid production—
notably that of butyrate, isobutyrate, and acetate—increased
with polydextrose ingestion. There were substantial changes in
fecal anaerobes after polydextrose intake. Bacteroides species
(B. fragilis, B.vulgatus, and B. intermedius) decreased,
whereas Lactobacillus and Bifidobacterium species increased.
The cecal mucosa whole-crypt labeling index increased, with
colonocyte proliferation mainly occurring in base compart-
ments, which provided an indirect confirmation of butyrate
production in the colon.
Conclusion: Polydextrose ingestion had significant dietary
fiber–like effects with no laxative problems. Am J Clin Nutr
2000;72:1503–9.
KEY WORDS Polydextrose, dietary fiber, blood glucose,
glycemic index, colonic fermentation, short-chain fatty acids, bowel
function, fecal anaerobes, Lactobacillus,Bifidobacterium, China
INTRODUCTION
Dietary fiber has attracted much attention since the 1970s
because of its beneficial effects on human physiology. Dietary
fiber was initially defined by Trowell and Burkitt as “remnants
of plant cell walls which were not hydrolyzed and digested by
human enzymes” (1). The definition recognized in China is “the
sum of food components that are not digested by intestinal
enzymes and absorbed into the body” (2). Dietary fiber is classi-
fied as water soluble or non–water soluble and includes cellu-
lose, gum, pectin, polysaccharides, and food additives.
Polydextrose is a polysaccharide synthesized by random poly-
merization of glucose, sorbitol, and a suitable acid catalyst at a
high temperature and partial vacuum. It is used widely in many
countries as a bulking agent and as a lower-energy ingredient
(4.2 kJ/g) in a variety of prepared foods. Polydextrose is not
digested or absorbed in the small intestine, and a large portion is
excreted in the feces (3). Several studies of polydextrose showed
physiologic effects consistent with those of dietary fiber (4–14).
Polydextrose is partially fermented in the large intestine, leading
to increased fecal bulk, reduced transit time, softer stools, and
lower fecal pH (4–9). Fermentation of polydextrose leads to the
growth of favorable microflora, diminished putrefactive micro-
flora, enhanced production of short-chain fatty acids (SCFAs),
and suppressed production of carcinogenic metabolites (eg, indole
and p-cresol) (3, 8, 10).
Therefore, the safety and efficacy of polydextrose as a water-
soluble bulking agent and fiber has been widely and thoroughly
investigated. The metabolic route has been established in ani-
mals and humans (3). The aim of the present trial was to evalu-
ate the effects of polydextrose on various body functions and
blood biochemical indexes in healthy Chinese subjects.
SUBJECTS AND METHODS
Subjects and polydextrose intake
One hundred twenty healthy volunteers participated in the
study, of whom 66 were men and 54 were women with average
ages of 32.9 and 29.4 y, respectively (Table 1). None of the
Am J Clin Nutr 2000;72:1503–9. Printed in USA. © 2000 American Society for Clinical Nutrition
Studies on the effects of polydextrose intake on physiologic
functions in Chinese people1–3
Zhong Jie, Luo Bang-yao, Xiang Ming-jie, Liu Hai-wei, Zhai Zu-kang, Wang Ting-song, and Stuart AS Craig
1503
1From the Departments of Gastroenterology, Endocrinology, Clinical
Bacteriology, and Emergency Medicine, Rui Jin Hospital, Shanghai, China;
the Medical Testing Center, Shanghai Second Medicine University, China;
and Danisco Cultor, Ardsley, NY.
2Supported by a grant from Pfizer (now Danisco Cultor) and organized
and supervised by the Shanghai Food Therapy Research Society.
3Address reprint requests to SAS Craig, Danisco Cultor, 440 Saw Mill
River Road, Ardsley, NY 10502. E-mail: stuart.craig@danisco.com.
Received February 18, 2000.
Accepted for publication May 21, 2000.
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
subjects reported fever or diarrhea during the trial period and all
subjects gave written consent. Subjects had no history of heart,
lung, kidney, liver, or metabolic disease. The subjects were
randomly assigned to 4 groups: A (0 g polydextrose/d; control),
B (4 g polydextrose/d), C (8 g polydextrose/d), and D (12 g poly-
dextrose/d). Polydextrose was provided in a double-blind man-
ner as a powder dissolved in warm water (100 mL) and was
drunk within 10 min of preparation. Polydextrose (Litesse) sam-
ples were provided by Pfizer (now Danisco Cultor), Ardsley, NY.
The study was conducted according to the ethical guidelines of
the Rui Jin Hospital Ethics Committee (Shanghai, China).
Study design
Before the feeding phase (days 4 to 7) began, subjects
underwent a screening that included a physical examination, a
dietary history, a family medical history, questions about bowel
habits, and measurements of clinical chemistry indexes. An oral-
glucose-tolerance test was conducted at baseline. On days 10
and 29, colonoscopy biopsy samples were taken from a cohort
group (n= 3/group) for pulse labeling of colonocytes with
[3H]thymidine to study colonic crypt cell proliferation. A dietary
control was conducted on days 4 to 1, ie, all meals were pro-
vided at the clinic (also on days 26–28). Before the feeding
phase began on day 1, baseline measurements of clinical
chemistry indexes and hemoglobin were made and feces were
collected for determination of pH, microflora, sterols, and mois-
ture (3-d pooled sample). The feeding phase, including polydex-
trose intake, began on day 0 and continued for 28 d. Body weight
was measured on day 0. Clinical chemistry indexes, hemoglobin,
body weights, and dietary records were conducted and a lifestyle
questionnaire was completed on days 7, 14, 21, and 28. Modified
glucose tolerance (ie, response to 50 g glucose + polydextrose)
and the glycemic index were determined on days –7, 0, and 29.
At the end of the feeding phase (day 28), feces were collected for
determination of pH, microflora, sterols, and moisture (3-d
pooled sample). An exit interview was conducted on day 29.
Feeding
All subjects were required to eat meals provided by the
clinic during the dietary control phase (days 4 to 1 and
26–28). The meals were prepared as typical Chinese food and
provided 9450–10710 kJ/d (2250–2550 kcal/d), 60–70 g pro-
tein/d, 50–60 g lipid/d, and 15–18 g fiber/d. Fruit consumption
was limited to one piece per day. During the feeding phase,
each subject was asked to record the foods consumed at each
meal, to approximate the amount of food consumed at each
meal, to report his or her daily activities, and to report any
adverse effects experienced after polydextrose intake. Weekly
visits to the clinical center were made to ensure compliance
and for the conduct of biochemistry tests.
Serum analysis
Plasma electrolytes, indexes of liver and renal function, fast-
ing blood sugar, lipids, and cholesterol were measured by using
a SMAC-II automatic biochemistry analyzer (Beckman, Palo
Alto, CA) in the clinical laboratories of Rui Jin Hospital. Gly-
cated hemoglobin (Hb A1c) was analyzed at the Shanghai
Endocrinology Institute with a DCA 2000 Analyzer (Bayer, Tar-
rytown, NY). Glucose tolerance and the glycemic index were
determined after ingestion of 50 g glucose. Blood samples were
taken to determine fasting glucose concentrations at baseline and
0, 30, 60, 90, 120, and 150 min after glucose consumption. The
blood glucose response to the modified oral-glucose-tolerance
test versus time was plotted for each subject. The glycemic index
of the test samples was calculated from the incremental area
under the curve (IAUC) of the blood glucose response divided by
the IAUC of the baseline response and expressed as a percentage
(15). Any area beneath the fasting concentration curve was
ignored. To determine the statistical significance of these differ-
ences, the glycemic index for each subject was calculated and
averaged within each group [groups A (control), B, C, and D]
and for each day (7, 0, and 29).
Stool sampling
The stool was collected over 3 consecutive days during the
dietary control period to determine fecal wet and dry weights
and pH. The fresh stool collected on days 1 and 28 was sent
to the Clinical Bacteriological Laboratory (Rui Jin Hospital)
within 1 h after defecation for culture of microflora and deter-
mination of SCFAs.
Fecal culture
Bacteria
The fresh stool samples were analyzed for the presence of
Bacteroides fragilis,B.vulgatus,B.intermedius, and Bifidobac-
terium and Lactobacillus species. Stool samples (0.5 g) were
diluted with water (2 mL) and drop-seeded onto anaerobic
blood agar plates. Plates were incubated for 48 h at 35C and
then counted, smeared, and Gram stained. An oxygen resistance
test was then performed, followed by incubation for 48 h at
30 C in selective culture media (Lactobacillus selection agar
for Lactobacillus, blood liver agar for Bifidobacterium). An
evaluation followed.
Short-chain fatty acids
Fecal SCFA determinations were made at the Medical Testing
Center of the Shanghai Second Medicine University. Samples of
0.5 g fresh feces were diluted with 2 mL normal saline solu-
tion followed by acidification with 1 mL of 50% H2SO4solution.
1504 JIE ET AL
TABLE 1
General characteristics of the subjects and polydextrose intakes
Group A (control) Group B Group C Group D
(n= 30) (n= 30) (n= 30) (n= 30)
Men/women 16/14 17/13 16/14 17/13
Average age (y) 30.2 32.7 31.7 29.6
Polydextrose intake (g/d) 0 4 8 12
Morning 0 4 4 6
Evening 0 0 4 6
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
This solution was then extracted with 2 mL ether and 1 L
extract was injected into the gas chromatograph. A GC-9A gas
chromatograph with a flame ionization detector (Shimadzu
Corp, Kyoto, Japan) was used. A column stationary phase of
10% fatty acids, a column temperature of 70C, and a detector
temperature of 230 C were used.
Measurement of colonocyte proliferation
During the pancolonoscopy, 3 biopsies of normal cecal
mucosa were taken from each subject. The samples were
immersed in Eagle’s medium and incubated for 3 h in an
equimolar sodium chloride solution. Next, the proliferating cells
were pulse labeled by incubating them with [3H]thymidine for
1 h. The samples were fixed in formalin, embedded in paraplast,
section-cut into 4-m slices, and stained with Schiff’s acid
reagent (Feulgen reaction). Sections were soaked in Ilford K2
emulsion (Ilford Ltd, Knutsford, United Kingdom) for 15 d by
using standard autoradiographic methods. The labeling fre-
quency of colonocytes was estimated by light microscopy in
15 longitudinally sectioned crypts of each run. The number of
labeled and unlabeled cells per crypt column was determined and
the whole-crypt labeling index (LI) (labeled cells per crypt col-
umn/labeled cells per crypt column + unlabeled cells per crypt
column ) was calculated. Each crypt was equally divided into
5 compartments, with compartment 1 representing the crypt base
and compartment 5 representing the crypt surface. Thus, the
compartment labeling index (CLI) (labeled cells in the compart-
ment/LI + unlabeled cells in the compartment) was calculated.
The mean LI or CLI values of 15 crypts were determined (16).
Observation of physiologic effects
Physiologic reactions after polydextrose intake were recorded,
including frequency of defecation, ease of defecation, abdominal
distention, abdominal cramps, diarrhea, hypoglycemic symptoms
(eg, sweating, pale skin, palpitation, and abdominal colic). Most
symptoms were rated on a scale of 1 to 10; ease of defecation
was rated on a scale of 3 to 3 and frequency of defecation was
reported as the number of times per day.
Statistics
The results are presented as means ±SDs. Analysis of vari-
ance was used to compare groups (before compared with after
polydextrose intake) and Dunnett’s multiple (pairwise) compari-
son procedure was used to determine differences between groups
A, B, C, and D. SAS (version 6.12; SAS Institute Inc, Cary, NC)
was used for the analyses.
RESULTS
Clinical biochemistry indexes (eg, measures of liver and renal
function, blood electrolytes, fasting blood sugar, triacylglycerol,
cholesterol, and serum Hb A1c) did not change significantly after
polydextrose intake for 28 d (data not shown). The IAUCs for
glucose were flattened in groups C and D after polydextrose
ingestion. The glycemic index decreased significantly from
baseline on days 0 and 28 in group D (Table 2).
The subjects who consumed polydextrose had marked changes
in bowel function (Tab l e 3 ). Groups B, C, and D showed signi-
ficant improvements in bowel function (increased frequency and
ease of defecation) and reported no laxation problems (abdomi-
nal distention, cramps, and diarrhea). There were no significant
differences between groups in ratings of abdominal distension
and no reports of abdominal cramps, diarrhea, hypoglycemic
symptoms, or other discomforts (data not shown). Fecal weights
(wet and dry) increased and fecal pH decreased after polydex-
trose intake (Table 4). This change was most significant in
groups C and D. The constitution and contents of SCFAs in
feces changed greatly after polydextrose intake (Tab l e 5 ). Par-
ticularly interesting was the significant increase in butyrate,
isobutyrate, and acetate in groups C and D. Significant changes
in fecal microflora were noted after polydextrose intake (Tab l e 6 ).
Bacteroides species decreased, whereas Lactobacillus and
POLYDEXTROSE AND PHYSIOLOGIC FUNCTION 1505
TABLE 3
Bowel function before and after polydextrose intake1
Group A (control) Group B Group C Group D
Before polydextrose intake
Frequency of defecation (times/d) 1.04 ±0.07 1.05 ±0.10 1.11 ±0.16 1.05 ±0.15
Ease of defecation 0.21 ±0.10 0.18 ±0.12 0.20 ±0.14 0.14 ±0.13
After polydextrose intake
Frequency of defecation (times/d) 1.10 ±0.21 1.47 ±0.282,3 1.74 ±0.372,3 1.89 ±0.262,3
Ease of defecation 0.41 ±0.15 1.36 ±0.222,3 1.88 ±0.262,3 2.35 ±0.302,3
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively. Ease of defecation was rated on a scale of –3 to 3.
2Significantly different from value before polydextrose intake within the same group, P< 0.01.
3Significantly different from the control group (baseline), P< 0.01.
TABLE 2
Glycemic indexes before and after polydextrose intake1
Day Group A (control) Group B Group C Group D
7 (baseline) 100 ±12 101 ±12 100 ±12 100 ±11
0 101 ±13 101 ±12 95 ±988±122,3
29 100 ±12 102 ±12 95 ±10 88 ±102,3
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively.
2Significantly different from baseline within the same group, P< 0.01.
3Significantly different from the control group (baseline), P< 0.01.
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
Bifidobacterium species increased in all groups after polydex-
trose intake. The whole-crypt CLI increased after polydextrose
ingestion (Table 7). Changes in the CLI occurred mainly in
groups C and D and in compartments 1, 2, and 3 (Table 8).
DISCUSSION
Polydextrose intake by the Chinese subjects led to many phys-
iologic effects associated with dietary fiber. The Chinese diet is
relatively low in fat (20% of energy from fat); therefore, we
did not expect to measure an effect of polydextrose on blood
lipids. Polydextrose had no influence on measured blood chem-
istry indexes. Fasting blood glucose and Hb A1c, indicators of
long-term stability of blood glucose concentrations, remained
unchanged. A polydextrose intake of 12 g (plus 50 g glucose)
flattened the postprandial glucose response significantly com-
pared with a 50-g glucose control. The glycemic indexes were
significantly lower in group D on day 0 (88 ±12%) and day 29
(88 ±10%) than at baseline (day 7). Note that the results on
day 0 indicate an immediate benefit from polydextrose, confirm-
ing that polydextrose is nonglycemic. This suggests that poly-
dextrose results in a reduction in glucose absorption from the
intestine, possibly related to delayed gastric emptying due to
polydextrose bulking and increased viscosity in the bowel
(17–19). Similar results were obtained by others (15, 20), who
found that the glycemic and insulin responses were markedly
flattened in healthy and diabetic volunteers after they consumed
fiber-enriched meals. Long-term consumption of foods high in
dietary fiber could reduce urinary glucose losses and improve the
control of diabetes.
Because of its excellent water-holding capacity, intake of
undigested polydextrose resulted in an increase in bowel peri-
stalsis and feces output. Most subjects reported a softening of
feces and improved ease of defecation after 2 d of polydex-
trose ingestion. There was a dose-response increase in the fre-
quency and ease of defecation and in both the wet and dry
weights of feces in groups B, C, and D after polydextrose intake.
The fecal wet weight increased by 25% and 40% in groups C
and D, respectively. There was a dose-response decrease in fecal
pH, due mainly to the production of SCFAs (21, 22). A high fecal
output and a low bowel pH can suppress the production of
enteric toxins, such as indole and p-cresol (8, 23, 24). This plays
an important role in the prevention of constipation and divertic-
ulosis and thereby reduces the risk of bowel cancer (25–27).
Abdominal distention, diarrhea, cramps, and hypoglycemic
symptoms were not reported by any of the subjects.
Dietary fiber is available for fermentation by anaerobes in
the colon. It can increase stool weight and change the constitu-
1506 JIE ET AL
TABLE 4
Stool weight and pH before and after polydextrose intake1
Group A (control) Group B Group C Group D
Before polydextrose intake
Fecal wet weight (g/d) 103 ±12.3 106 ±16.4 101 ±13.6 98 ±12.5
Fecal dry weight (g/d) 32.2 ±8.3 34.0 ±6.8 31.5 ±7.4 29.6 ±8.9
Fecal pH 7.06 ±0.18 7.05 ±0.21 7.00 ±0.25 6.93 ±0.18
After polydextrose intake
Fecal wet weight (g/d) 106 ±15.9 115 ±16.7 128 ±27.42,3 142 ±18.33,4
Fecal dry weight (g/d) 34.5 ±8.8 38.3 ±13.1 41.8 ±16.3247.8 ±18.23,4
Fecal pH 7.04 ±0.18 6.89 ±0.1656.71 ±0.2336.37 ±0.273,4
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively.
2,4 Significantly different from value before polydextrose intake within the same group: 2P< 0.05, 4P< 0.01.
3,5 Significantly different from the control group (baseline): 3P< 0.01, 5P< 0.05.
TABLE 5
Fecal short-chain fatty acid contents before and after polydextrose intake1
Group A (control) Group B Group C Group D
mg/g stool
Before polydextrose intake
Acetate 4.04 ±0.16 4.12 ±0.24 4.13 ±0.21 4.09 ±0.22
Propionate 1.48 ±0.17 1.48 ±0.23 1.52 ±0.25 1.52 ±0.27
Butyrate 0.92 ±0.16 0.95 ±0.21 0.98 ±0.17 0.96 ±0.20
Isobutyrate 0.17 ±0.06 0.18 ±0.08 0.20 ±0.10 0.19 ±0.09
Isovalerate 0.41 ±0.13 0.39 ±0.12 0.43 ±0.18 0.42 ±0.15
After polydextrose intake
Acetate 4.12 ±0.19 4.18 ±0.26 4.70 ±0.332,3 5.12 ±0.313,4
Propionate 1.50 ±0.24 1.52 ±0.22 1.55 ±0.36 1.48 ±0.35
Butyrate 0.94 ±0.23 1.10 ±0.17 1.31 ±0.242,3 1.41 ±0.343,4
Isobutyrate 0.20 ±0.15 0.23 ±0.12 0.26 ±0.1820.29 ±0.164,5
Isovalerate 0.38 ±0.15 0.36 ±0.13 0.40 ±0.26 0.41 ±0.28
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively.
2,4 Significantly different from value before polydextrose intake within the same group: 2P< 0.05, 4P< 0.01.
3,5 Significantly different from the control group (baseline): 3P< 0.01, 5P< 0.05.
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
tion of microflora, eg, increase the Lactobacillus content (28).
Lactic acid, produced by Lactobacillus, can reduce the intesti-
nal pH. Bifidobacterium has a strong inhibiting effect on
Escherichia coli and Bacteroides in an acidic environment. The
mechanism is probably related to the production of antibiotic-
like substances during the proliferation of some specific strains
(28). In the present study, Lactobacillus and Bifidobacterium
concentrations were significantly higher and Bacteroides
species were significantly lower in fresh stool after all poly-
dextrose intakes. The relation between the proliferation of Bifi-
dobacterium species with an acidic environment, dietary fiber
sources, and amounts of Bifidobacterium in the bowel require
further investigation.
The main products of fermentation are hydrogen and carbon
dioxide gases and SCFAs—primarily acetate, propionate, and
butyrate (26, 29). The relative amount of each product depends
POLYDEXTROSE AND PHYSIOLOGIC FUNCTION 1507
TABLE 6
Fecal microflora contents before and after polydextrose intake1
Group A (control) Group B Group C Group D
109/g stool
Before polydextrose intake
Bacteroides fragilis 1.35 ±0.46 1.54 ±0.37 1.37 ±0.40 1.36 ±0.39
Bacteroides vulgatus 0.78 ±0.26 0.84 ±0.28 0.66 ±0.22 0.80 ±0.31
Bacteroides intermedius 0.34 ±0.12 0.46 ±0.1150.37 ±0.24 0.33 ±0.16
Lactobacillus 0.24 ±0.15 0.26 ±0.08 0.32 ±0.14 0.28 ±0.12
Bifidobacterium 0.57 ±0.23 0.46 ±0.21 0.40 ±0.1630.52 ±0.19
After polydextrose intake
Bacteroides fragilis 1.53 ±0.52 0.58 ±0.322,3 0.24 ±0.072,3 0.27 ±0.082,3
Bacteroides vulgatus 0.83 ±0.26 0.45 ±0.193,4 0.13 ±0.052,3 0.16 ±0.042,3
Bacteroides intermedius 0.35 ±0.18 0.26 ±0.174,5 0.09 ±0.032,3 Not detected2,3
Lactobacillus 0.27 ±0.10 1.10 ±0. 282,3 1.36 ±0.532,3 1.92 ±0.482,3
Bifidobacterium 0.48 ±0.17 1.54 ±0.372,3 3.05 ±1.132,3 5.29 ±1.742,3
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively.
2,4 Significantly different from value before polydextrose intake within the same group: 2P< 0.01, 4P< 0.05.
3,5 Significantly different from the control group (baseline): 3P< 0.05, 5P< 0.01.
TABLE 7
Whole-crypt labeling index before and after polydextrose intake1
Group A (control) Group B Group C Group D
Before 0.0664 ±0.006 0.0673 ±0.008 0.0675 ±0.005 0.0680 ±0.008
After 0.0671 ±0.007 0.0816 ±0.0162,3 0.1021 ±0.0173,4 0.1313 ±0.0153,4
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively.
2,4 Significantly different from value before polydextrose intake within the same group: 2P< 0.05, 4P< 0.01.
3Significantly different from the control group (baseline), P< 0.01.
TABLE 8
Compartment labeling index before and after polydextrose intake
Compartment
1234 5
Group A (control)
Before 0.102 ±0.012 0.136 ±0.025 0.068 ±0.014 0.040 ±0.006 0.007 ±0.002
After 0.105 ±0.011 0.133 ±0.022 0.072 ±0.016 0.038 ±0.003 0.008 ±0.003
Group B
Before 0.106 ±0.020 0.137 ±0.030 0.071 ±0.016 0.042 ±0.008 0.008 ±0.002
After 0.118 ±0.034 0.148 ±0.035 0.087 ±0.027 0.051 ±0.01620.007 ±0.004
Group C
Before 0.105 ±0.011 0.134 ±0.029 0.070 ±0.015 0.039 ±0.005 0.007 ±0.003
After 0.138 ±0.0323,4 0.228 ±0.0494,5 0.146 ±0.0314,5 0.058 ±0.0173,4 0.009 ±0.005
Group D
Before 0.107 ±0.016 0.135 ±0.034 0.073 ±0.020 0.035 ±0.008 0.006 ±0.002
After 0.177 ±0.0283,4 0.283 ±0.0544,5 0.161 ±0.0344,5 0.057 ±0.0244,5 0.010 ±0.003
1x
–±SD. 0, 4, 8, and 12 g polydextrose/d for 28 d in groups A, B, C, and D, respectively.
2,4 Significantly different from the control group (baseline): 2P< 0.05, 4P< 0.01.
3,5 Significantly different from value before polydextrose intake within the same group: 3P< 0.05, 5P< 0.01.
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
mainly on the substrates entering the colon and the types of
microflora that proliferate (30). Acetate is produced in the
largest amounts from fermentation of dietary fiber. It is absorbed
into the bloodstream together with propionate and is metabolized
in the liver and peripheral tissues. Butyrate, arguably the most
important fermentation product, is generally regarded as an
energy resource for colonocytes. Butyrate has been shown to
have desirable effects on colonic epithelial cells, including stabi-
lization of DNA and down-regulation of oncogenes (31–33). The
production of acetate and butyrate in the feces of subjects who
ingested 8 or 12 g polydextrose (groups C and D, respectively)
increased significantly. Thus, a polydextrose intake ≥8 g/d can
result in substantial production of butyrate and consequent desir-
able effects on the human colon.
Measurement of colonic crypt cell proliferation provides an
indirect measure of SFCA production in the colon, particularly
butyrate (16, 34–37). The present study showed that consumption
of polydextrose promoted the growth of normal cecal epithelial
cells. The whole-crypt labeling index increased after all polydex-
trose intakes, especially in groups C and D. Growth occurred
mainly in the base compartments of the crypt, ie, compartments
1–3. Scheppach et al (16) found similar results after incubating
normal human cecal colonocytes directly with SCFAs; the effect
was most pronounced for butyrate and propionate. These investi-
gators also found growth to be significant only in compartments
1–3. Thus, the results of Scheppach et al agree with those of the
present study (Table 5). The effects of butyrate on regulation of
cell phase and on morphologic changes remain unknown.
In conclusion, polydextrose is a dietary fiber that has many
physiologic benefits. Consumption of polydextrose significantly
improved bowel function, softened the feces, and improved the
ease of defecation, with no adverse effects. Polydextrose intake
inhibited excessive glucose absorption from the small intestine
and was fermented in the lower gut to produce SCFAs, including
butyrate. Polydextrose promoted the proliferation of favorable
intestinal microflora and decreased the pH of the bowel. There-
fore, daily intake of ≥4–12 g polydextrose improves physiologic
function without adverse effects.
We acknowledge the assistance of Hank Frier (formerly with Pfizer, now
with Slim Fast) and John Troup (formerly with Pfizer, now with Novartis) for
helping design the study and Gary Williams, Seth Thompson, and Bandaru
Reddy (American Health Foundation) for reviewing and providing sugges-
tions about the manuscript.
REFERENCES
1. Trowell HC, Burkitt DP. Concluding considerations. In: Burkitt DP,
Trowell HC, eds. Refined carbohydrate foods and disease. Some
indication of dietary fiber. New York:Academic Press, 1975:333–45.
2. Jian-xian Z. Active polysaccharides. In: Functional food. Beijing:
China Light Industry Publishing House, 1995:10.
3. Pfizer Inc. Polydextrose food additive petition. New York: Pfizer Inc,
1978. (FDA petition 9A3441.)
4. Hamanaka M. Polydextrose as a water-soluble dietary fiber.
Syokuhin Kogyo 1987;9:73–80.
5. Tomlin J, Read NW. A comparative study of the effect on colon
function by feeding isphagula husk and polydextrose. Aliment Phar-
macol Ther 1988;2:513–9.
6. Nakagawa Y, Okamatsu H, Fujii Y. Effects of polydextrose feeding
on the frequency and feeling of defecation in healthy female volun-
teers. J Jpn Soc Nutr Food Sci 1990;43:95–101.
7. Oku T, Fujii Y, Okamatsu H. Polydextrose as dietary fiber: hydroly-
sis by digestive enzyme and its effect on gastrointestinal transit time
in rats. J Clin Biochem Nutr 1991;11:31–40.
8. Endo K, Kumemura M, Nakamura K, et al. Effect of high choles-
terol diet and polydextrose supplementation on the microflora, bac-
terial enzyme activity, putrefactive products, volatile fatty acid
(VFA) profile, weight, and pH of the feces in healthy volunteers.
Bifidobact Microflora 1991;10:53–64.
9. Achour L, Flourie B, Briet F, Pellier P, Marteau P, Rambaud J. Gas-
trointestinal effects and energy value of polydextrose in healthy
nonobese men. Am J Clin Nutr 1994;59:1362–8.
10. Wang X, Gibson GR. Effects of the in vitro fermentation of
oligofructose and inulin by bacteria growing in the human large
intestine. J Appl Bacteriol 1993;75:373–80.
11. Harada E, Hashimoto Y, Syuto B. Polydextrose induces precocious
cessation of intestinal macromolecular transmission and develop-
ment of digestive enzymes in the suckling rat. Comp Biochem Phys-
iol 1995;111A:479–85.
12. Figdor SK, Rennhard HH. Caloric utilization and disposition of
14C-polydextrose in the rat. J Agric Food Chem 1981;29:1181–9.
13. Figdor SK, Bianchine JR. Caloric utilization and disposition of
14C-polydextrose in man. J Agric Food Chem 1983;31:389–93.
14. Beereboom JJ. Third annual workshop conference on foods, nutri-
tion and dental health. Chicago: American Dental Association
Health Foundation Research Institute, 1979:10–2.
15. Jenkins DJA, Jenkins AL. Dietary fiber and the glycemic response.
Proc Soc Exp Biol Med 1985;180:422–31.
16. Scheppach W, Bartram P, Richter A, Richter F, Liepold W, Kasper H.
Enhancement of colonic crypt proliferation in man by short-chain fatty
acids (SCFA). In: Southgate DAT, Waldron K, Johnson IT, Fenwick
GR, eds. Dietary fiber: chemical and biological aspects. Cambridge,
United Kingdom: The Royal Society of Chemistry, 1991:233–7.
17. Holt S, Heading RC, Carter DC, Prescott LF, Tothill P. Effect of a
gel fibre on gastric emptying and absorption of glucose and parac-
etamol. Lancet 1979;1:636–9.
18. Morgan LM, Goulder TJ, et al. The effect of unabsorbable carbohy-
drate on gut hormones: modification of post-prandial GIP secretion
by guar. Diabetologia 1979;17:85–9.
19. O’Connor N, Tredger J, Morgan L. Viscosity differences between
various guar gums. Diabetologia 1981;20:612–5.
20. Rivellese A, Riccardi G, Giacco A, et al. Effect of dietary fibre on
glucose control and serum lipoproteins in diabetic patients. Lancet
1980;2:9957–9.
21. Muir JG, Lu ZX, Young GP, Cameron-Smith D, Collier GR, O’Dea
K. Resistant starch in the diet increases breath hydrogen and serum
acetate in human subjects. Am J Clin Nutr 1995;61:792–9.
22. Fleming SE, Marthinsen D, Kuhnlein H. Colonic function and fer-
mentation in men consuming high fiber diets. J Nutr 1983;113:
2535–44.
23. Phillips J, Muir JG, Birkett A, et al. Effect of resistant starch on
fecal bulk and fermentation-dependent events in humans. Am J Clin
Nutr 1995;62:121–30.
24. Van Dokkum W, Pikaar NA. Physiological effects of fibre-rich types
of bread. Dietary fibre from bread: digestibility by the intestinal
microflora and water-holding capacity in the colon of human sub-
jects. Br J Nutr 1983;50:61–74.
25. Cummings JH, Englyst HN. Fermentation in the human large intes-
tine and the available substrates. Am J Clin Nutr 1987;45:1243–55.
26. Gustafsson BE. The physiological importance of the colonic
microflora. Scand J Gastroenterol Suppl 1982;77:1171–31.
27. Cummings JH, Bingham SA. Dietary fibre, fermentation and large
bowel cancer. Cancer Surv 1987;6:601–21.
28. Mortensen PB, Hove H, Clausen MR, Holtug K. Fermentations to short
chain fatty acids and lactate in human faecal batch cultures. Intra- and
inter-individual variations versus variations caused by changes in fer-
mented saccharides. Scand J Gastroenterol 1991; 26:1285–94.
1508 JIE ET AL
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
29. Cummings JH, Hill MJ, Bine ES, Branch WJ, Jenkins DJ. The effect
of meat protein and dietary fibre on colonic function and metabo-
lism. II. Bacterial metabolites in feces and urine. Am J Clin Nutr
1979;32:2094–101.
30. Macfarlane GT, Cummings JH. The colonic flora, fermenta-
tion, and large bowel digestive function. In: Phillips SF, Pem-
berton JH, Shorter RG, ed. The large intestine: physiology,
pathophysiology and disease. New York: Raven Press, 1991:
51–92.
31. Newmark HL, Lupton JR. Determinants and consequences of
colonic luminal pH: implications for colon cancer. Nutr Cancer
1990;14:161–73.
32. Cummings JH. Colonic absorption: the importance of SCFAs in
man. Scand J Gastroenterol Suppl 1984;93:117–31.
33. Whitehead RH, Young GP, Bhathal PS. Effects of short chain fatty
acids on a new human colon carcinoma cell line (LIM1215). Gut
1986;27:1457–63.
34. Jacobs LR, Lupton JR. Effects of dietary fibres on rat large bowel
mucosal growth and cell proliferation. Am J Physiol 1994;246:
G378–85.
35. Stephen AM, Cummings JH. Mechanism of action of dietary fibre
in the human colon. Nature 1980;284:293–4.
36. Malhotra SL. Fecal urobilinogen levels and pH of stools in popula-
tion groups with different incidence of cancer of the colon, and their
possible role in its etiology. J R Soc Med 1982;75:709–14.
37. McIntyre A, Young GP, Taranto T, Gibson PR, Ward PB. Different
fibers have different regional effects of rat colon. Gastroenterology
1991;101:1274–81.
POLYDEXTROSE AND PHYSIOLOGIC FUNCTION 1509
Downloaded from https://academic.oup.com/ajcn/article-abstract/72/6/1503/4729522
by guest
on 24 April 2018
