ORIGINAL ARTICLE

Acacia Gum is a Bifidogenic Dietary Fibre with High

Digestive Tolerance in Healthy Humans
Christine Cherbut1, Catherine Michel1, Virginie Raison1, Thierry Kravtchenko2 and
Meance Severine2
From the 1Human Nutrition Research Centre, INRA, BP 71627, FR-44316 Nantes cedex 03, France and the 2Colloides
Naturels International, CNI, BP 4151, 76723 Rouen cedex 3, France
Correspondence to: C. Cherbut, INRA, BP 71627, FR-44316 Nantes cedex 03, France. Fax: 33-2-40-67-50-12; E-mail:
[email protected]

Microbial Ecology in Health and Disease 2003; 15: 43
/50

The objective of the present study was to determine whether acacia gum (GUM) is a prebiotic fibre and to evaluate its intestinal tolerance in
healthy subjects. The effects of increasing doses of GUM were compared to those of sucrose (SUC) on stool output, concentration of the
main bacterial populations in stools, and occurrence and severity of intestinal symptoms (flatulence, bloating, abdominal cramps, diarrhoea).
Ingestion of GUM 10 and 15 g/day for 10 days increased total lactic acid-producing bacteria and bifidobacteria counts in stools, without
affecting total anaerobe and aerobe counts. The magnitude of this selective effect was greater in subjects with a low initial faecal
concentration of bifidobacteria. Faecal digestibility of GUM was around 95% and its caloric value was estimated to range between 5.5 and
7.7 kJ/g. In addition, stool weight increased 30% because of greater faecal water content. Digestive tolerance of GUM was high and not
statistically different from that of SUC up to a daily dose of 30 g. Above this dose, the main complaint was excessive flatulence. However, the
mean degree of severity remained mild ( B/1), even at doses /50 g/day. Other intestinal events were rarely reported. Thus GUM is a very well
tolerated dietary fibre with bifidogenic properties believed to benefit intestinal health. Key words: acacia gum, prebiotic, soluble fibre, stool
output, digestive tolerance.

INTRODUCTION
Nutritionists agree that increasing dietary fibre intake up to
25
/30 g/day is part of a healthy diet. Some dietary fibres
influence digestion processes and thus metabolic responses,
while others improve transit and stool output (1). In
addition, some of them appear to be essential for mucosal
barrier integrity and intestinal ecosystem balance by
selectively stimulating the growth and/or activity of bacteria, such as bifidobacteria or lactobacilli, which is
considered to benefit the host (2). However, some fibres
are rapidly fermented into short-chain fatty acids and
associated with a pronounced increase in gas production,
a condition that may induce undesirable symptoms: bloating, flatus, abdominal cramps and diarrhoea. Thus,
although dietary fibres have health-promoting effects, these
adverse events may limit their consumption.
Acacia gum (GUM) or gum arabic is a soluble dietary
fibre obtained from the stems and branches of Acacia
senegal and A. seyal . It is composed mainly of complex
polysaccharides (95%) that consist of highly branched
galactan polymers, with galactose and/or arabinose side
chains, possibly terminated by rhamnose or glucuronic acid
residues (3). While /80% of current production is used by
the food industry for various applications (emulsification,
# Taylor & Francis 2003. ISSN 0891-060X

encapsulation, coating, gum candies, etc.), GUM is traditionally consumed by African and Indian populations to
improve digestive comfort and intestinal transit. In vitro, its
fermentation is slow (4) and supports the growth of
bifidobacteria (5). Thus, GUM may be a bifidogenic soluble
dietary fibre that would not induce uncomfortable intestinal
side effects in the healthy consumer. The purpose of the
present study was to investigate the effects of GUM on
bowel performance and the bacterial composition of stools
and to evaluate the occurrence and severity of digestive
symptoms associated with increasing intake of GUM in
healthy subjects.

SUBJECTS AND METHODS

Substrates
GUM (Fibregum ; Colloids Naturels International,
Rouen, France) consists of branched arabino-galactans
polymers (92.8% of the dry matter) containing the following
sugars (molar percentage in parentheses): galactose (38),
arabinose (46), rhamnose (4), glucuronic acid (6.5), methylglucuronic acid (5.5). Its molecular weight is 882 kDa and
its polydispersity is 1.74. The effects of GUM were
compared with that of sucrose (SUC, placebo), and shortchain fructo-oligosaccharides (FOS; Actilight , EridaniaMicrobial Ecology in Health and Disease
DOI: 10.1080/08910600310014377

44

C. Cherbut et al.

Beghin-Say, Paris, France), a rapidly fermented soluble

fibre. The three substrates were provided by CNI as
odourless, white powders packed in 5-g sachets.
Subjects
Twenty subjects (9 men and 11 women, 22
/38 years of age)
were included in the study after undergoing a preliminary
medical check-up. None had a history of gastrointestinal
disease or antibiotic or laxative use during the preceding 3
months. All were lactose absorbers, as tested by an H2
breath test after ingestion of 50 g of lactose in water. They
gave their informed consent to the protocol, which was
approved by the local ethics committee (CCPPRB, Nantes,
France). Each subject ingested GUM, SUC and FOS
powders that were identical in appearance and colour.
The powders were added to orange juice to mask different
tastes, and ingested two to six times daily at the subjects
convenience.
Experimental design
The study was divided into two independent experiments.
The first was intended to assess the effects of GUM on stool
output and faecal bacterial concentration in 10 volunteers
(4 men and 6 women, aged 22
/33 years). The second sought
to determine the nature and severity of digestive symptoms
during consumption of increasing doses of GUM in 10
volunteers (5 men and 5 women, aged 22
/38 years). The
two experiments were separated by at least a 2-month
interval.
Experiment 1. Faecal output and the bacterial composition of stools were determined over four 10-day periods (no
interruptions) in which supplements of SUC 10 g/day,
GUM 10 g/day, SUC 15 g/day and GUM 15 g/day were
given successively. Menus were established for each day of
each period to keep the diet as constant as possible,
especially for fibre intake (15
/20 g/day). Foods containing
inulin, oligosaccharides or alcohol sugars were excluded
from the diet, as well as dairy products enriched with
bifidobacteria or lactobacilli.
Volunteers were asked to record the times at which daily
meals were consumed and sachets ingested, as well as the
number of sachets used. All stools excreted on days 5
/10 of
each period were collected in plastic boxes and maintained
at 48C under anaerobic conditions using an AnaerocultA
(Merck, Darmstatd, Germany) for rapid transportation
(within 1 h) to the laboratory. Each stool was weighed
and homogenized before one 1-g aliquot was collected for
immediate bacterial analysis and two others for determination of dry matter (16 h at 1038C) and total neutral sugars
content. The analysis of the faecal microflora was made by
V.R., who knew the order of the successive periods of
supplementation. The other analyses were made blindly.
Thus, only a part of the experiment was double-blinded in
nature.

Experiment 2. Tolerance to increasing amounts of GUM

(10
/70 g/day) was compared to that of SUC and FOS
ingested according to the same escalation schedule lasting
18 days: 10 g/day during days 1 and 2, then a 5 g/day
increase every 2 days up to 30 g/day on days 9 and 10,
followed by a further 10 g/day increase every 2 days up to 70
g/day on days 17 and 18. Intake periods for each substrate
were separated by at least 2 weeks of wash-out. The
experiment was designed as a double-blind, randomized
cross-over study.
Throughout experiment 2, volunteers were asked to avoid
foods known to promote abdominal symptoms (cabbage,
artichoke, onions, beans, lentils, wheat bran, prunes and
plum juice) and to record the time at which meals were
consumed, and their composition, and the times at which
sachets were ingested, and the number used. They also filled
out a diary card concerning tolerance and gastrointestinal
symptoms. The following items were recorded: number of
bowel movements per day, stool consistency (1 /liquid;
2 /soft or 3 /hard) and size (1/small, 2 /medium or
3 /large), episodes of flatulence and abdominal discomfort
(bloating, borborygmi, cramps) and uneasiness (nausea,
headaches, others). Subjects rated symptom severity as
absent (0), mild (1), moderate (2), or severe (3). Volunteers
were asked to report diarrhoea (one or more watery stools
or more than three liquid stools per day) immediately, as
well as any grade 3 symptoms.
At day 6 (15 g/day) and 12 (30 g/day) of each period, H2
excretion was measured in breath. Daily intake of the tested
substrate was provided by thirds in meals at breakfast,
lunch and dinner. A latex balloon was used to collect
samples of expired gas before breakfast and then every hour
for 16 h after breakfast. Each sample was stored in a syringe
before being analysed for H2 by gas chromatography
(Quintron microlyser model DP, Quintron Instruments
Co., MI, USA). On the 2 days during which breath tests
were performed, meals were standardized and special care
was taken to ensure identical timing for meals and substrate
ingestion.

In vitro fermentation
In vitro fermentation of GUM was compared to that of
FOS, chosen as a rapidly fermented substrate, and sugar
beet fibre (Fibrex , Fibrex AB, Malmo, Sweden), chosen as
a slowly fermented substrate. The study was carried out
according to Barry et al. (6), with slight modifications.
Briefly, two faecal slurries were prepared by homogenizing
faeces (20% w/v) from two healthy volunteers in anaerobic
CO2-saturated nutritive buffer and then filtered through six
layers of surgical gauze. Volunteers, who ate Western-style
diets, had not taken antibiotics for at least 3 months before
the study. Serum bottles (60 ml working volume) containing
10 ml of substrate solution (2% w/v in carbonate/phosphate
buffer) were inoculated with 10 ml of faecal slurry. Bottles

Intestinal tolerance of acacia gum in humans

with no additional substrates served as controls. Air was

displaced by a flow of oxygen-free nitrogen. Cultures were
incubated at 378C for 24 h under anaerobic conditions. Gas
production was recorded every hour for 8 h, then at 12 and
24 h, and samples were taken at 0 and 24 h for pH
determination, SCFA analysis and substrate disappearance.
All incubations were performed in duplicate and the study
was repeated twice.
Analyses
Microbiological analysis. All bacterial analyses were
performed within 2 h of defecation. Faecal samples (1 g)
were serially diluted 10-fold in half-strength peptone water
anaerobic broth. Triplicate plates for each dilution were
then inoculated in appropriate agars and incubated at 378C
either aerobically or in an anaerobic chamber (H2 5%, CO2
10%, N2 85%). Total anaerobic bacteria were counted on
Wilkins
/Chalgren agar, facultative anaerobes on nutrient
agar, Bacteroides fragilis group on modified Bacteroides
mineral salt, lactobacilli on Rogosa agar, and total lactic
acid-producing bacteria on Man
/Rogosa
/Sharpe agar.
Bifidobacteria were counted on Beerens medium after
microscopic examination of gram-stained bacteria. Clostridia were determined on Wilkins
/Chalgren medium after
heat treatment at 808C for 10 min. All media were
purchased from Oxoid (Basingstoke, UK).
Short-chain fatty acids. A mercuric chloride (1% w/v) and
orthophosphoric acid (5% w/v) solution was added (10%) to
aliquots of fermentation media. The samples obtained were
then stored at /208C. Before analysis, samples were thawed
and centrifuged. A methyl valerate solution (86.1 mM) was
then added (10%) to supernates as internal standard. SCFA
were measured by gas chromatographic procedures (7).
In vitro substrate disappearance. The amount of total
sugars was determined in the initial substrate/inoculum
solutions and their residues after fermentation for 24 h.
Faecal neutral sugars. The faecal neutral sugars were
determined in each stool collected for 5 days from each
subject, throughout the SUC 15 g/day and GUM 15 g/day
periods of experiment 1, by gas-liquid chromatography after
the stools had been hydrolysed at 358C for 1 h in H2SO4
(720 g/kg) (8).
Calculations and statistical analysis
All the data are given as means9/SEM. All statistical
analyses were performed with Statview 5.0 software
(SAS Institute).
Experiment 1. The effect of the substrate and dose on
stool output was assessed first by two-way and then by oneway ANOVA. The bulking index of GUM, expressed in
grams of wet stool per gram of ingested GUM, was
calculated by subtracting the wet stool weight measured
on day i (i /5 to 10) of the SUC period from that measured
on the corresponding day of the gum period, and then

45

dividing the difference by the daily dose of GUM. Faecal

digestibility of the total neutral sugars of GUM was
calculated after correction for the excretion of these sugars
during the control period.
The effect of the substrate and dose on faecal bacteria
concentrations was assessed by two-way and then by oneway ANOVA. The influence of the basal concentration of
bifidobacteria on the magnitude of the bifidogenic effect of
GUM was evaluated. On the basis of their mean faecal
concentration of bifidobacteria during the SUC ingestion
period, volunteers were classified into two groups of five
subjects: one with /9.5 log cfu/g of stool and the other
with less. The difference between the bifidobacterial concentration (expressed in cfu/g) measured on day i of the
SUC period and that measured on the corresponding day of
the GUM period was then calculated for each day and each
volunteer. The means of the two groups were compared by
the Mann
/Whitney test for unpaired data.
Experiment 2. The occurrence dose was regarded as the
first dose producing a symptom that the subject constantly
rated higher with GUM or FOS than with SUC. If some of
the symptoms were not experienced with the maximal dose
(70 g/day), it was assumed that these symptoms would have
occurred at the next step dose (80 g/day). Mean occurrence
doses for GUM and FOS were compared for each symptom
using the Wilcoxon rank test for paired data. At each dose,
the percentage of subjects who experienced symptoms at a
level constantly higher with GUM or FOS than with SUC
was calculated. The 50% effective dose (ED50) was then
plotted for each symptom.
The effect of each substrate dose on symptom severity
was determined using one-way ANOVA followed by a post
hoc test. The possible correlation between the severity of
each symptom and the doses of ingested substrates was
studied by linear regression analysis.
Hydrogen excretion (expressed in ppm /16 h) was
quantified using a trapezoidal method to compute the
area under the discontinuous curve of breath H2 concentration. The effect of the substrate and dose on mean H2
excretion was determined first by two-way and then by oneway ANOVA.
In vitro fermentation. SCFA production (mmol/l) was
calculated after subtraction of the amount of SCFA present
in the initial inocula. Substrate disappearance was expressed as percentage of the initial substrate amounts:
(IS/FS)/100/IS, where IS and FS are the initial and final
quantities of total sugars, respectively.
RESULTS
Stool output
The number of stools per week was slightly, although not
significantly, increased by ingestion of GUM (10 and 15 g/
day) as compared with SUC. Accordingly, daily stool
weight was higher during GUM consumption (Table I).

46

C. Cherbut et al.

Table I
Stool output measured in subjects (n/10) consuming supplements of sucrose (SUC) or acacia gum (GUM) added to a controlled diet
Supplement (g/day)

Number of stools/7 days

Wet stool weight (g/day)

Stool water (g/day)

Stool dry matter content (%)

SUC 10
GUM 10
SUC 15
GUM 15

7.79/0.7
8.19/0.6
7.89/0.2
8.09/0.6

103.69/8.3
115.89/10.4
100.09/8.6
130.29/11.4 *

77.69/6.4
84.69/8.0
71.99/7.3
108.49/9.2 *

26.19/1.4
28.29/1.7
27.39/2.5
25.59/1.8 *

Data are means9/SEM.

* Different from SUC 15 g/day, p B/0.05.

GUM intake also increased water excretion and decreased

the dry matter content of stools, although these effects were
only significant at a dose of 15 g/day. The overall bulking
index for GUM was 1.69/0.5 g of added stool per gram of
ingested GUM. Faecal output of total neutral sugars
tended to be slightly higher during the ingestion of 15 g/
day GUM (3.199/0.21 g/day) as compared with the period
of 15 g/day SUC (2.79/0.21 g/day, p /0.107). Faecal
digestibility of the neutral sugars of GUM, as calculated
after correction for the excretion of these sugars during the
SUC period, was 96.59/2.0% (range 87.7
/100).
Faecal microflora
After substrate administration, the number of total bacteria
was constant throughout the study, although differences
were recorded in the counts of certain bacterial groups. A
significant increase in bifidobacteria, lactobacilli and total
lactic acid-producing bacteria was observed during GUM
consumption (Table II). Considerable variations in faecal
bifidobacteria concentrations occurred among subjects
during the first period of SUC ingestion (10 g/day).
Bifidobacteria counts ranged from 7.5 to 10.2 log cfu/g in
this period. These inter-individual variations were reduced
during the second SUC period following the period of
GUM 10 g/day (Table II). A significant interaction was
found between the subject and the substrate for faecal

bifidobacteria concentration (p/ 0.007), which led us to

compare the magnitude of the GUM-induced increase in
bifidobacteria in subjects with low ( B/9.5 log cfu/g) and
high ( /9.5 log cfu/g) basal concentration. The ability of
GUM to stimulate bifidobacteria growth was significantly
higher in subjects whose basal faecal concentration was B/
9.5 log cfu/g (Fig. 1).
Digestive symptoms
None of the tested substrates induced major symptoms of
digestive intolerance. Diarrhoea was observed in 2 of the 20
subjects with FOS 50 g/day and in one subject with 60 g/day.
The mean degree of severity of the different symptoms did
not reach a score of 1 (mild), regardless of the substrate
involved (Table III and Fig. 2). For the overall study,
symptoms were more frequent with FOS than GUM
(19.59/5.0 and 9.19/2.5 symptoms per subject, respectively),
and both of these substrates induced more frequent
symptoms than SUC (3.39/1.2 symptoms per subject).
The degree of symptom severity reported during consumption of substrates was higher with FOS than GUM, and
higher with both fermentable substrates than with SUC.
However, digestive problems were not significantly different
with GUM and SUC below a dose of 30 g/day (Fig. 2). The
mean occurrence doses for flatus, bloating and borborygmi
were lower with GUM than FOS (Table IV). Accordingly,

Table II
Mean viable counts and pH of five faecal samples collected in subjects (n/10) consuming supplements of sucrose (SUC) or acacia gum
(GUM) added to a controlled diet
Bacteria log cfu/g wet stool

SUC 10 g/day GUM 10 g/day SUC 15 g/day GUM 15 g/day Substrate

p

Total aerobes
7.949/0.11
Total anaerobes
10.459/0.05
Bifidobacteria
9.139/0.16
Bacteroides
8.979/0.07
Clostridia
5.439/0.09
Lactobacilli
8.659/0.18
Total lactic acid-producing bac- 9.349/0.16
teria
Faecal pH
6.749/0.07

Dose
p

Substrate/dose
p

7.819/0.09
10.429/0.04
9.759/0.14 *
8.869/0.16
5.439/0.11
9.139/0.18 *
10.119/0.06 *

7.729/0.13
10.499/0.05
9.589/0.06
9.109/0.07
5.469/0.11
8.759/0.14
9.759/0.06

7.559/0.13
10.509/0.05
10.029/0.07 $
9.249/0.13
5.239/0.10
9.079/0.17
10.229/0.05 $

0.19
0.89
B/0.0001
0.85
0.29
0.02
B/0.0001

0.04
0.42
0.002
0.03
0.42
0.91
0.006

0.89
0.72
0.44
0.29
0.26
0.65
0.11

6.659/0.07

6.649/0.07

6.619/0.08

0.42

0.34

0.72

Data are means9/SEM.

* Statistically different from SUC 10 g/day, p B/0.05.
$ Statistically different from SUC 15 g/day, p B/0.05.

Intestinal tolerance of acacia gum in humans

Fig. 1. Increase induced by acacia gum consumption (I 10 g/day

and j 15 g/day) in faecal bifidobacteria in subjects with low ( B/9.5
log cfu/g) and high ( /9.5 log cfu/g) basal concentration. *High vs
low basal concentration were statistically different, p B/0.02. $Gum
15 g/day vs 10 g/day were statistically different, p/0.05.

the 50% effective dose was higher with GUM than FOS for
flatus and bloating. Thus, a daily dose of 17 g of FOS
induced flatus in half of the subjects, whereas 46 g/day of
GUM were needed to induce the same symptoms (Table
IV).
Breath tests
Consumption of GUM and FOS, as compared with SUC,
raised breath hydrogen concentration (p B/0.001). The
quantity of H2 excreted in breath, as calculated by the
area under the 16-h excretion curve, was higher with FOS
than with GUM (Fig. 3), and the effect depended on the
ingested dose (p /0.06).
In vitro fermentation
The volume of gas produced during in vitro incubation of
faecal bacteria with carbohydrates reflects the extent of
fermentation, and the kinetics of gas production is indicative of the fermentation rate. Gas volume increased steadily
during the first 12 h of GUM incubation then remained
stable, but rose very rapidly (within 2 h) to its maximal

47

Fig. 2. Flatus and bloating severity reported by healthy subjects

consuming sucrose (I) or acacia gum (j) or fructo-oligosaccharides (k) at doses increasing from 10 to 70 g/day. Symptom severity
was rated as absent (0), mild (1), moderate (2), or severe (3). *GUM
is statistically different from sucrose, p B/0.05. $FOS is statistically
different from sucrose.

value during FOS incubation and then decreased slightly

before stabilizing after 8 h. Gas volume increased more
progressively during sugar beet fibre fermentation (Fig. 4).
The final pH of the in vitro incubation medium was lower
with FOS (4.99/0.1) than with GUM (5.79/0.3) and sugar
beet fibre (6.39/0.2). GUM led to a total SCFA concentration (47.89/9.4 mmol/l) between those induced by FOS
(61.69/2.4 mmol/l) and sugar beet fibre fermentation
(39.59/5.1 mmol/l). Accordingly, the disappearance of total
sugars was higher after the 24-h fermentation of FOS
(76.89/15.7%) than after fermentation of GUM (68.49/
24.2%) and sugar beet fibre (56.69/18.9%).

DISCUSSION
The present study shows that acacia gum (GUM) did not
induce adverse gastrointestinal effects even when consumed
at high doses, and demonstrates that the gum is a
bifidogenic factor. Dietary fibre may cause several symptoms of digestive discomfort during the first days of intake
and progressive adaptation is generally recommended. Our
study showed that the digestive tolerance of GUM was
high. Flatus, bloating and borborygmi occurred at lower

Table III
Mean degree of symptom severity and correlation with the dose during regular consumption of sucrose (SUC), acacia gum (GUM) or fructooligosaccharides (FOS)
Symptom

SUC
Degree

Flatus
Bloating
Borborygmi
Cramps

0.159/0.04
0.069/0.03
0.019/0.01
0.019/0.01

GUM
r

p
0.536
0.118
0.312
0.612

Data are means9/SEM.

* Statistically different from SUC, p B/0.05.
$ Statistically different from GUM, p B/0.05.

Degree
0.170
0.780
0.451
0.107

0.619/0.09 *
0.159/0.04 *
0.109/0.04 *
0.229/0.07 *

FOS
r

p
0.842
0.749
0.711
0.134

Degree
0.009
0.032
0.048
0.751

0.929/0.09 *$
0.409/0.07 *$
0.419/0.08 *$
0.299/0.08 *

p
0.587
0.391
0.733
0.079

0.126
0.338
0.038
0.853

48

C. Cherbut et al.

Table IV
Mean occurrence doses and 50% effective doses (ED50) for each
symptom during the regular consumption of acacia gum (GUM) and
fructo-oligosaccharides (FOS)
Symptom

Flatus
Bloating
Borborygmi
Cramps

Occurrence dose (g/day)

ED50 (g/day)

GUM

FOS

GUM

FOS

53.59/2.5
679/2.1
70.59/2.2
719/1.9

26.59/2.1 *
509/3.1 *
51.59/3.3 *
67.89/2.1

46
65.7
60.7
55

17.5
41.7
60
57.5

doses with FOS than with GUM. The effect of FOS was
very similar to that described previously (9, 10). Both of
these studies found that tolerance did not improve during
12 or 15 days of regular intake. The main complaint related
to all doses of FOS was excessive flatulence, but daily doses
B/20 g/day resulted in only minor complaints, while doses
/50 g/day resulted in abdominal cramps and diarrhoea (9).
In the present study, GUM did not induce flatulence below
the dose of 30 g/day and daily doses higher than 50 g/day

Fig. 3. Cumulated excretion of breath hydrogen for 16 h after

ingestion of sucrose (SUC), acacia gum (GUM) or fructooligosaccharide (FOS) supplements (I 15 g or j 30 g), added to
standardized meals.

Fig. 4. Gas volume collected during in vitro incubation of a human

faecal inoculum alone (/





); and added with sugar beet fiber ( */



*/), fructo-oligosaccharides ( */) or acacia gum (-----).

did not provoke any abdominal cramps or diarrhoea. The

recommended daily intake for dietary fibre worldwide
ranges between 25 and 40 g/day, and it is expected that
GUM intake may account for 0
/50% of these doses (i.e. 0
/
20 g/day), which is largely below the 50% effective doses
inducing abdominal symptoms. The better tolerance of
GUM as compared with that of FOS may be due to several
factors. First, probably owing to its high molecular weight,
GUM was fermented at a slower rate than FOS and led to
higher pH after incubation for 24 h. The lower pH recorded
after fermentation of FOS in comparison with GUM may
result from the higher production of SCFA as well as an
accumulation of lactate. We did not monitor it in the
present study; however, we previously observed that lactate
accumulated only during the in vitro fermentation of FOS
and not during that of GUM (5). We assume that this effect
may arise because bacterial glycolysis of FOS to lactate is
more rapid than the further metabolization of this intermediate to SCFA. Acid pH or lactic acid may excite some
visceral sensory nervous fibres and lower nociceptive
threshold for intestinal distension (11). Secondly, breath
hydrogen excretion was lower during GUM than FOS
intake, possibly because of lower gas production. Even if
intestinal gas tolerance is normally high, some apparently
normal subjects can retain gas and develop abdominal
distension and symptoms (12). The origin of the increase in
gas production after the FOS intake is not obvious, as this
sugar is preferentially fermented by bifidobacteria and
lactic acid-producing bacteria, which do not yield any gas.
One possible explanation is that the lactate produced by
these bacteria can be further utilized by other bacterial
groups which are gas producers (13). Another explanation
could be that, despite their selectivity towards lactic acid
bacteria, FOS might also be metabolized to a lesser degree
by gas-forming organisms such as clostridia (14).
In addition to its high level of intestinal tolerance, GUM
increased stool weight. Its bulking capacity is similar to that
of pectins and guar gum (15), as well as inulin and FOS
(16). This effect was associated with a slight rise in dry
matter excretion and with a major rise in water excretion.
As faecal digestibility of GUM was high, the increase in
stool dry matter and water content was probably due to
faecal bacteria, as previously shown in chronic renal failure
patients (17). It has already been reported that gum arabic
completely disappears during its passage through the
gastrointestinal tract both in animals (18, 19) and in
humans (20, 21). Our results confirm that the neutral
sugars from GUM are extensively degraded in the human
intestine and indicate that the overall digestibility of GUM
could be around 95% in healthy humans consuming the
GUM for at least 5 days. A lower percentage of disappearance was found after in vitro fermentation for 24 h. As
there are no enzymes that are able to hydrolyse GUM
polysaccharides in the human stomach and small intestine,

Intestinal tolerance of acacia gum in humans

it is unlikely that the discrepancy comes from the digestion

of GUM in the upper digestive system. Furthermore, many
results from animal models showed that gum arabic resists
digestion in the small intestine and enters into the caecum
(18, 19, 22). A more plausible explanation is that the
difference between the fermentability in vitro and in vivo
was due to the prior adaptation of colonic microflora in the
human trial. The need for such an adaptation for a
complete degradation of gum arabic has been demonstrated
in rats (19, 22) and strongly suggested in humans (21).
Because of this necessary adaptation of the flora to
thoroughly metabolize GUM, the digestible energy value
of GUM may vary. Nevertheless, because energy is lost to
faeces, predominantly as unfermented carbohydrates and as
bacterial biomass, the digestible energy value of GUM is
less than the 17.5 kJ/g value for available starch. According
to the correlations between fermentability and digestible
energy value of non-starch polysaccharides, or between the
metabolizable energy of fibre and production of SCFA in
vitro, found in two European inter-laboratory studies (6,
23), the caloric value of GUM could range between 5.5 and
7.7 kJ/g, thus B/2 kcal/g and close to the 1.5 kcal/g value
proposed for nutritional labelling (24).
For the first time, this study demonstrated that GUM
increased the counts and the proportion of bifidobacteria in
human stools and is thus a strong candidate prebiotic. In
vitro studies showed that GUM supported the growth of
pure cultures of bifidobacteria strains (25) and increased
total lactic acid-producing bacteria counts in continuous
cultures of human faecal microflora (5). In the stools of one
volunteer, Wyatt et al. (21) measured an increase in the
proportion of bacteria able to ferment the gum, among
which species of Bacteroides and Bifidobacterium were
found. These preliminary findings suggested that GUM
could be prebiotic; however, its effect in healthy humans was
not established. Our results show that, compared with SUC,
GUM intake (10 and 15 g/day) raised faecal concentrations
of total lactic acid-producing bacteria and bifidobacteria,
without changing total anaerobe and aerobe counts. Moreover, bacteroides and clostridial groups were not affected by
GUM, indicating that the effect was specific. The increase
in the proportion of lactic acid-producing bacteria and
bifidobacteria may be related to the acidification of colonic
contents during GUM fermentation. A decrease in pH
could have promoted the growth of acid-tolerant bacteria,
such as lactobacilli and bifidobacteria. However, the pH of
colonic contents is decreased by many fermentable fibres,
most of which do not appear to be bifidogenic. Therefore,
even though acidification may favour the bifidogenic effect
of GUM, it is likely that in our study this effect was mainly
due to the ability of the gum to support the growth of
Bifidobacterium spp., as reported for B. longum and B.
adolescentis (25). The fact that GUM contains high
amounts of galactose could account for this property.

49

Several works comparing different carbohydrates have

reported that galactose-containing oligosaccharides were
the most effective in stimulating growth of bifidobacteria
(26
/28). In addition, we observed that bifidobacteria
concentration was particularly increased in subjects whose
initial concentration was low. Such an inverse relationship
has been observed previously for other prebiotics (28, 29).
Therefore, GUM, like other prebiotics, may be particularly
useful for targeting certain groups of individuals, such as
the elderly who have diminished bifidobacteria populations.
In conclusion, this study suggests that GUM is a
prebiotic fibre, as it was able to selectively raise the
proportions of lactic acid bacteria and bifidobacteria in
healthy subjects. It was fermented slowly, produced shortchain fatty acids, and its faecal digestibility was around
95%. As in the case of other soluble dietary fibres, it
increased stool output by augmenting the water content of
stools. In addition, its intestinal tolerance appeared to be
excellent and high daily doses could be consumed without
inducing any adverse intestinal events. Thanks to all these
properties, GUM may improve intestinal well-being and
health, although this potential remains to be confirmed in
further nutritional trials including a large number of
volunteers.

ACKNOWLEDGEMENTS
We thank B. Flourie, I. Rowland, R. Walker and T. Sakata for their
useful advice regarding the experimental design and expression of
the results.

REFERENCES
1. Spiller GA. CRC Handbook of Dietary Fiber in Human
Nutrition, 3rd edn. Boca Raton, FL: CRC Press, 2001: 709.
2. Gibson GR, Roberfroid MB. Dietary modulation of the human
colonic microflora: introducing the concept of prebiotics. J
Nutr 1995; 125: 1401
/12.
3. Islam AM, Phillips GO, Slijivo A, Snowden MJ, Williams PA.
A review of recent developments on regulatory, structural and
functional aspects of gum arabic. Food Hydrocol 1997; 11:
357
/65.
4. Titgemeyer EC, Bourquin LD, Fahey GC, Garleb KA.
Fermentability of various fiber sources by human fecal bacteria
in vitro. Am J Clin Nutr 1991; 53: 1418
/24.
5. Michel C, Kravtchenko TP, David A, Gueneau S, Kozlowski F,
Cherbut C. In vitro prebiotic effects of Acacia gums onto the
human intestinal microbiota depends on both botanical origin
and environmental pH. Anaerobes 1998; 4: 257
/66.
6. Barry JL, Hoebler C, Macfarlane GT, Macfarlane S, Mathers
JC, Reed KA, Mortensen PB, Nordgaard I, Rowland IR,
Rumney CJ. Estimation of the fermentability of dietary fibre in
vitro: a European interlaboratory study. Br J Nutr 1995; 74:
303
/22.
7. Jouany JP. Volatile fatty acid and alcohol determination in
digestive contents, silage juices, bacterial cultures and anaerobic fermentor contents. Sci Alim 1982; 2: 131
/44.
8. Hoebler C, Barry JL, David A, Delort-Laval J. Rapid acid
hydrolysis of plant cell wall polysaccharides and simplified
quantitative determination of their neutral monosaccharides by

50

9.

10.

11.

12.
13.

14.

15.

16.

17.

18.

C. Cherbut et al.
gas
/liquid chromatography. J Agric Food Chem 2001; 37:
360
/7.
Briet F, Achour L, Flourie B, Beaugerie L, Pellier P,
Franchisseur C, Bornet F, Rambaud JC. Symptomatic response
to varying levels of fructo-oligosaccharides consumed occasionally or regularly. Eur J Clin Nutr 1995; 49: 501
/7.
Stone-Dorshow T, Levitt MD. Gaseous response to ingestion
of a poorly absorbed fructo-oligosaccharide sweetener. Am J
Clin Nutr 1987; 46: 61
/5.
Su X, Gebhart GF. Mechanosensitive pelvic nerve afferent
fibers innervating the colon of the rat are polymodal in
character. J Neurophysiol 1998; 80: 2632
/44.
Serra J, Azpiroz F, Malagelada JR. Intestinal gas dynamics and
tolerance in humans. Gastroenterology 1998; 115: 542
/50.
Bernalier A, Dore J, Durand M. Biochemistry of fermentation.
In: Gibson GR, Roberfroid MB, eds. The Colonic Microbiota:
Nutrition and Health. London: Kluwer Academic, 1999: 27
/
35.
Probert HM, Gibson GR. Investigating the prebiotic and gasgenerating effects of selected carbohydrates on the human
colonic microflora. Lett Appl Microbiol 2002; 35: 473
/80.
Cummings JH. The effect of dietary fiber on fecal weight and
composition. In: Spiller GA, ed. Handbook of Dietary Fiber in
Human Nutrition, 2nd edn. Boca Raton, FL: CRC Press, 1993:
263
/349.
Gibson GR, Beatty ER, Wang X, Cummings JH. Selective
stimulation of bifidobacteria in the human colon by oligofructose and inulin. Gastroenterology 1995; 108: 975
/82.
Bliss DZ, Stein TP, Schleifer CR, Settle RG. Supplementation
with gum arabic fiber increases fecal nitrogen excretion and
lowers serum urea nitrogen concentration in chronic renal
failure patients consuming a low-protein diet. Am J Clin Nutr
1996; 63: 392
/8.
McLean Ross AH, Eastwood MA, Brydon WG, Busuttil A,
McKay LF. A study of the effects of dietary gum arabic in the
rat. Br J Nutr 1984; 51: 47
/54.

19. Walter DJ, Eastwood MA, Brydon WG, Elton RA. An

experimental design to study colonic fibre fermentation in the
rat: the duration of feeding. Br J Nutr 1986; 55: 465
/79.
20. McLean Ross AH, Eastwood MA, Brydon WG, Anderson JR,
Anderson DMW. A study of the effects of dietary gum arabic in
humans. Am J Clin Nutr 1983; 37: 368
/75.
21. Wyatt GM, Bayliss CE, Holcroft JD. A change in human faecal
flora in response to inclusion of gum arabic in the diet. Br J
Nutr 1986; 55: 261
/6.
22. Tulung B, Remesy C, Demigne C. Specific effects of guar gum
or gum arabic on adaptation of cecal digestion to high fiber
diets in the rat. J Nutr 1987; 117: 1556
/61.
23. Livesey G, Smith T, Eggum BO, Tetens IH, Nyman M,
Roberfroid M, Delzenne N, Schweizer TF, Decombaz J.
Determination of digestible energy values and fermentabilities
of dietary fibre supplements: a European interlaboratory study
in vivo. Br J Nutr 1995; 74: 289
/302.
24. Phillips GO. Acacia gum (gum arabic): a nutritional fibre;
metabolism and calorific value. Food Add Contam 1998; 15:
251
/64.
25. Crociani F, Alessandrini A, Mucci MM, Biavati B. Degradation of complex carbohydrates by Bifidobacterium spp. Int J
Food Microbiol 1994; 24: 199
/210.
26. Jaskari J, Kontula P, Siitonen A, Jousimies-Somer H, MattilaSandholm T, Poutanen K. Oat -glucan and xylan hydrolysates
as selective substrates for Bifidobacterium and Lactobacillus
strains. Appl Microbiol Biotechnol 1998; 49: 175
/81.
27. Djouzi Z, Andrieux C. Compared effects of three oligosaccharides on metabolism of intestinal microflora in rats inoculated
with a human faecal flora. Br J Nutr 1997; 78: 313
/24.
28. Rycroft CE, Jones MR, Gibson GR, Rastall RA. A comparative in vitro evaluation of the fermentation properties of
prebiotic oligosaccharides. J Appl Microbiol 2001; 91: 878
/87.
29. Roberfroid MB, Van Loo JAE, Gibson GR. The bifidogenic
nature of chicory inulin and its hydrolysis products. J Nutr
1998; 128: 1
/9.