The Effect of a Multispecies Probiotic on the IntestinalMicrobiota and Bowel Movements in Healthy VolunteersTaking the Antibiotic AmoxycillinCatherina J.M. Koning, M.Sc.,1 Daisy M.A.E. Jonkers, Ph.D.,1,2 Ellen E. Stobberingh, Ph.D.,2Linda Mulder, M.Sc.,3 Frans M. Rombouts, Ph.D.,4 and Reinhold W. Stockbr¨ugger, M.D., Ph.D.11Division of Gastroenterology-Hepatology and 2Department of Medical Microbiology, University HospitalMaastricht, Maastricht, The Netherlands; 3Winclove Bio Industries B.V., Amsterdam, The Netherlands; and4Laboratory of Food Microbiology, Department of Agrotechnology and Food Sciences, Wageningen University,Wageningen, The Netherlands
One of the side effects of antimicrobial therapy is a disturbance of the intestinal microbiotapotentially resulting in antibiotic-associated diarrhea (AAD). In this placebo-controlled double-blindstudy, the effect of a multispecies probiotic on the composition and metabolic activity of theintestinal microbiota and bowel habits was studied in healthy volunteers taking amoxycillin.
Forty-one healthy volunteers were given 500 mg amoxycillin twice daily for 7 days and wererandomized to either 5 g of a multispecies probiotic, Ecologic
daily for 14 days. Feces and questionnaires were collected on day 0, 7, 14, and 63. Feces wasanalyzed as to the composition of the intestinal microbiota, and β-glucosidase activity, endotoxinconcentration, Clostridium difficile toxin A, short chain fatty acids (SCFAs), and pH were determined. Bowel movements were scored according to the Bristol stool form scale.
Mean number of enterococci increased significantly from log 4.1 at day 0 to log 5.8 (day 7) and log6.9 (day 14) cfu/g feces (P < 0.05) during probiotic intake. Although no other significant differenceswere observed between both intervention groups, within each group significant changes were foundover time in both microbial composition and metabolic activity. Moreover, bowel movements with afrequency ≥3 per day for at least 2 days and/or a consistency ≥5 for at least 2 days were reportedless frequently in the probiotic compared to the placebo group (48% vs 79%, P < 0.05).
Apart from an increase in enterococci no significant differences in microbial composition andmetabolic activity were observed in the probiotic compared with the placebo group. However,changes over time were present in both groups, which differed significantly between the probioticand the placebo arm, suggesting that the amoxycillin effect was modulated by probiotic intake. Moreover, the intake of a multispecies probiotic significantly reduced diarrhea-like bowelmovements in healthy volunteers receiving amoxycillin. INTRODUCTION
the cases of AAD are thought to be due to a disturbance ofthe intestinal microbiota by antibiotics, which is associated
One of the collateral effects of antimicrobial therapy is
with loss of colonization resistance (leading to overgrowth
antibiotic-associated diarrhea (AAD), which can occur
of potential pathogens), changes in carbohydrate digestion
shortly after antibiotic intake up to 8 wk after cessation (1, 2).
and production of short-chain fatty acids (SCFAs), altered
The incidence of AAD ranges from 5–39%, depending on the
metabolism of bile acids, and changes in both the mucosal
definition of diarrhea, the type of antibiotic used, and host fac-
and systemic immune response (7). In addition, antibiotics
tors (2). In general, amoxycillin, amoxycillin/clavulanic acid,
may have direct allergic and toxic effects on the mucosa, di-
clindamycin, and cephalosporines are associated with a high
rect effects on immune-cell function, and pharmacological
risk of AAD (3). AAD may range from mild disturbances to
effects on intestinal motility (7–9). Possible consequences
severe pseudomembranous colitis due to Clostridium difficile
of AAD in health-care facilities include an increase in the
(4). This bacterium is thought to be the causative agent in up
incidence of nosocomial infections and an increase in mor-
to 20% of AAD patients; however, the mechanisms causing
bidity and mortality, longer hospitalization, and higher costs
the majority of cases of AAD are not clear (5, 6). Most of
of care (2). Although in general practice AAD is often merely
Koning et al.
considered a nuisance, it may lead to a lack of compliance of
The objective of this placebo-controlled, randomized,
antibiotic intake, which is associated with the development
double-blind study was to evaluate the effect of a multi-
of antibiotic resistance (10). Furthermore, antibiotic use and
species probiotic on the composition of the intestinal mi-
the subsequent disturbance of the intestinal microbiota is a
crobiota in healthy volunteers during and after amoxycillin
risk factor for the development of irritable bowel syndrome
intake. Moreover, the effect of the probiotic on the metabolic
activity of the intestinal microbiota and on bowel habits was
Probiotics, which are defined as “mono- or mixed cultures
of live microorganisms that, when applied to animal or hu-man, beneficially affect the host by improving the propertiesof the indigenous microbiota,” may prevent and restore an
imbalance caused by antibiotics and are therefore of increas-ing interest for the prevention and treatment of AAD (13). Subjects
Several probiotic strains have been used in controlled stud-
Healthy volunteers between 18 and 65 yr of age were eligible
ies, aiming at the prevention and treatment of AAD, such
for the study. Exclusion criteria were: smoking, pregnancy,
as Lactobacillus acidophilus, Lactobacillus rhamnosus GG,
lactation, hypersensitivity to β-lactam antibiotics or tetracy-
Bifidobacterium longum, Enterococcus faecium, and Saccha-
cline, pre-existing bowel pathology (including irritable bowel
romyces boulardii, and resulted in a significant decrease in
syndrome, inflammatory bowel disease, diverticulitis, and
the incidence of diarrhea (0–10% in the probiotic versus 14–
cancer), treatment with immune-suppressive medication or
27% in the placebo group) (14–22). However, other studies
immune-compromised subjects, diarrhea or constipation (in
failed to show any benefit from probiotics in the prevention
the last 3 days prior to inclusion), allergic and inflammatory
of AAD (23–25). Two meta-analyses on the use of probiotics
reactions, as well as infections within 2 wk prior to inclu-
in the prevention of AAD evaluated nine and seven placebo-
sion. Furthermore, the volunteers were not allowed to use:
controlled, double-blind trials, and reported an odds ratio of
(a) gastric acid inhibitors, laxatives, antidiarrhea medication,
0.37 and a relative risk of 0.40, respectively, in favor of pro-
or antibiotics for at least 2 months before the start of and dur-
biotic administration (26, 27). A recent meta-analysis, which
ing the study; (b) corticosteroids for at least 4 wk before the
included 25 randomized controlled trials, confirmed these
start of and during the study; (c) other probiotics and prebi-
findings (relative risk of 0.43), and showed that the probi-
otics for at least 2 wk before the start of and during the study.
otic efficacy could be attributed to three types of probiotics:
Finally, participants were asked to continue their ordinary di-
S. boulardii, L. rhamnosus GG, and probiotic mixtures (28).
etary habits. All volunteers gave written informed consent.
Most studies on probiotics and AAD have only investigated
The study was approved by the medical ethics committee of
the development of diarrhea (i.e., clinical outcome); only in
the University Hospital Maastricht, The Netherlands.
a few studies the effect on the composition of the fecal mi-crobiota was also examined, albeit to a very limited degree
Study Design
The study was executed according to a parallel, randomized,
It has recently been demonstrated that multispecies and, to
placebo-controlled, double-blind design. The total duration
a lesser extent, multistrain probiotics have certain advantages
of the intervention and follow-up period was 63 days. Vol-
over monostrain preparations. Mixed preparations may com-
unteers received 500 mg amoxycillin twice daily from day
plement each other’s effect through synergism and/or sym-
1–7 and were randomized to receive either 5 g of a multi-
biosis (31). Ouwehand et al. reported, for example, that the
species probiotic or 5 g placebo twice daily from day 1–14. in vitro adhesion of B. lactis Bb12 was more than doubled
This resulted in three time periods defined as: day 1–7, “the
by the presence of L. rhamnosus GG and L. delbrueckii spp.
antibiotic/probiotic period,” in which all volunteers received
Bulgaricus (32). A progressive increase in B. lactis growth
amoxycillin in combination with either probiotic or placebo;
and acidification in the presence of L. acidophilus in vitro
day 8–14, “the probiotic only period,” in which volunteers
was demonstrated by Gomes et al. (33).
received either probiotic or placebo; day 15–63, “the post-
It is well known that the composition of the fecal mi-
treatment follow-up period.” Amoxycillin was taken with
crobiota and its role in colonization resistance, but also
milk before breakfast and dinner, while the placebo or pro-
its metabolic activity (producing several bacterial enzymes,
biotic was taken before lunch and before bedtime. The time
short chain fatty acids [SCFAs], amines, and bacteriocins),
between antibiotic and probiotic intake had to be at least 2 h.
will affect the host. However, few studies have assessed the
Fresh fecal samples were collected on day 0 (i.e., baseline), 7,
influence of probiotics on the metabolic activity of the intesti-
14, and 63. On the same day, a questionnaire was filled out in-
nal microbiota during and after antibiotic treatment. Com-
cluding questions on bowel movements (stool frequency and
monly, studies on AAD concerned the use of monospecies
consistency [ranging from 1 = hard lumps to 7 = completely
probiotics. The results of these studies are difficult to gener-
watery] according to the Bristol stool form scale (34)), use of
alize, since there is a lot of variation between species; proper-
pre- and probiotics, other medication taken, (drastic) change
ties that apply to one strain are not necessarily applicable to
of eating habits, and compliance. In addition, a short ques-
tionnaire on bowel habits and side effects (nausea, abdominal
Effect of a Multispecies Probiotic on the Intestinal Microbiota and Bowel Movements
cramps, bloating, flatulence, or other) had to be completed
emulsion (36). Before inoculation of the egg yolk-neomycin
daily during probiotic/placebo intake.
agar, fecal dilutions were heated at 80◦C for 10 min.
Blood agar and eosin-methylene blue agar plates were in-
Probiotic
cubated aerobically at 37◦C for 24 h. Sabouraud GM+C and
The multispecies probiotic (Ecologic AAD) and the placebo
KF-streptococcus agar plates were incubated aerobically at
were kindly provided by Winclove Bio Industries, Amster-
37◦C and 42◦C, respectively, for 48 h. Fastidious anaerobic,
dam, The Netherlands. Ecologic AAD consists of 10 differ-
bile-esculine, egg yolk, and LAMVAB agar plates were in-
ent bacterial species at each 108 colony forming units (cfu)/g
cubated under anaerobic conditions at 37◦C for 48 h.
the total dose being 109 cfu/g (B. bifidum W23, B. lactisW18, B. longum W51, E. faecium W54, L. acidophilus W37
Viability of E. faecium W54
and W55, L. paracasei W72, L. plantarum W62, L. rham-
From each KF-streptococcus agar plate two dominant
nosus W71, and L. salivarius W24), 5% mineral mix (potas-
colonies were isolated on both day 7 and 14 and purified
sium chloride [∼67.3%], magnesium sulphate [∼32.6%] and
on blood agar plates. Isolates were frozen at −80◦C awaiting
manganese sulphate [∼0.1%]) and 15% Raftilose synergy1
further identification. At the end of the study, the enterococci
(inulin enriched with oligofructose). Each participant con-
of 10 individuals (i.e., 40 colonies), who had received probi-
otic treatment, were typed by pulse field gel electrophoreses
twice daily for 2 wk. Sachets were dissolved in lukewarm
(PFGE), using SmaI according to the method described by
water, left for 10 min, stirred, and thereafter ingested. The
van den Braak et al. (37, 38).
placebo sachets were indistinguishable in color, smell, andtaste from the probiotic sachets but contained no bacteria. Bacterial Enzyme Activity Bacterial β-glucosidase activity was determined as previ- Sample Processing
ously described (39). Briefly, fecal dilutions were mixed (1:1)
Fecal samples were brought to the laboratory within 12 h after
with 0.1 M PBS (pH 6.8), sonicated for 1 min, and centrifuged
defecation and divided into three portions: (a) ten grams was
at 1,700 g for 15 min. The supernatants were lyophilized for
centrifuged at 47,000 g for 2 h at 4◦C to obtain fecal water,
75 min by Speed-Vac (Savant DNA 120, GMI, Inc., Ram-
which was frozen immediately in 2-fold at −80◦C for anal-
sey, MN) and the remaining fractions were used to determine
ysis of endotoxin concentrations and determination of pH;
β-glucosidase (at 420 nm) activity by using p-nitrophenyl
(b) five grams was diluted (1:4) with peptone water (Oxoid
β-D-glucopyranoside as a substrate.
CM9, Basingstoke, Hants, U.K.) supplemented with cysteine(2.1 mM) and glycerol (30%). Bacterial cultures of the fe-
Endotoxin
cal dilution were performed immediately and the remainder
The endotoxin (i.e., lipopolysaccharide) concentration was
was frozen at −20◦C for the subsequent analyses of enzyme
determined in fecal water using the Limulus amoebocyte
activities, Clostridium difficile toxin A, and SCFAs, (c) the
lysate endochrome technique (Endosafe, end point chro-
remaining fecal sample (1–15 grams) was frozen directly at
mogenic analysis endochrome test kit, Charles River, Kent,
U.K.). The analysis was performed according to the man-ufacturer’s specifications under pyrogen-free conditions. Bacteriological Culture
Pyrogen-free water was used to dilute the fecal samples, and
Tenfold serial dilutions of the fecal dilution were made
the test-solutions and as negative control. The detection range
in physiological saline (0.85%) with cysteine-HCl (0.05%)
of the assay was 0.015 to 0.12 EU/mL (9 EU/ng). Concen-
and 40 µL of these dilutions was inoculated using a spiral
tration of fecal endotoxin was expressed as nanogram of en-
plater (Eddy Jet v1.2, IUL-instruments, Barcelona, Spain)
onto the following agar plates: blood agar (Oxoid CM271)for total (facultative) aerobic bacteria, eosin-methylene blue
Clostridium difficile Toxin A
(methylthioninium chloride) agar (Oxoid CM69) for enter-
Clostridium difficile toxin A was determined using an
obacteriaceae, KF-streptococcus agar (Oxoid CM701) for en-
enzyme-linked fluorescent immunoassay technique (VI-
terococci, fastidious anaerobic agar (Laboratory M LabGo,
C. difficile Toxin A II assay, bioMerieux, Lyon,
Lancashire, U.K.) for total (facultative) anaerobic bacteria,
France). The analysis was performed according to the
bile-esculine agar (Becton Dickinson 287920, La Pont de
manufacturer’s specifications using the VIDAS system
Claix, France) for Bacteriodes spp., LAMVAB agar for lac-
tobacilli, Sabouraud agar with gentamicin and chlorampheni-col (GM+C) (Becton Dickinson 254041) for yeasts and egg
Short-Chain Fatty Acids
yolk-neomycin agar for spore-forming clostridia. LAMVAB
SCFAs were measured in the fecal dilutions using gas-
agar was prepared according to the method described by
liquid chromatography. The gas-liquid chromatography sys-
Hartemink et al. (35). Egg yolk-neomycin agar was prepared
tem consisted of a CP9002 gas chromatograph equipped
by adding a sterile neomycin solution (final concentration,
with a flame ionization detector in conjunction with Mae-
100 µg/mL) to egg yolk agar with freshly prepared egg yolk
stro software (Chrompack, Middelburg, The Netherlands) for
Koning et al.
calculations. The chromatographic column used was WCOT
Based on data from previous probiotic studies, it was esti-
fused silica (25 m × 0.32 mm id), coated with FFAP-CB df
mated that 19 volunteers per treatment group would provide
0.3. This column was used in an isothermal mode at 140◦C
80% power to detect a one log difference in numbers of spe-
and both the injector and detector temperature were 270◦C.
cific microorganisms cultured, assuming a variance of 1.1 and
The sample size was 1.0 µL, which was split 50:1 to give a
a 2-sided significance level of 0.05.
0.02-µL sample on the column. Helium was used as the car-rier gas with a head pressure of 0.8 bar. SCFAs were extractedand analyzed as previously described (39). Subjects pH
Forty healthy volunteers completed the study, 19 in the pro-
The pH of fecal water was determined using a PHM standard
biotic (5 men and 14 women, mean age 25.5, SD 10.2 yr)
pH meter with a PHC3006 electrode (Radiometer Nederland
and 21 in the placebo group (10 men and 11 women, mean
age 28.2, SD 11.5 yr). One subject in the probiotic group wasfound to be allergic to amoxycillin and had to be excluded. Protein Concentration
Two subjects in the placebo group did not complete the daily
Homogenized fecal samples were diluted (1:99) in 0.1 M
questionnaire. On day 14, subjects delivered fecal samples to
PBS (pH 6.8, 5–7◦C) and added to BioRad Assay Protein
the hospital and subsequently handed in the daily question-
Dye Reagent (1:1). After 30 min, the absorbance was read at
naire. As a consequence information from the questionnaire
595 nm. Concentrations of proteins were calculated from a
was available for day 1–13. The compliance for antibiotic
standard curve for proteins ranging from 0–120 µg/mL and
intake was at least 93%, and for probiotic/placebo intake at
expressed as mg total protein per gram feces.
least 97% in both groups. One subject in the placebo groupand three in the probiotic group incidentally (i.e., maximally
Defecation Score
twice a week) consumed yogurt containing L. rhamnosus GG
In this study a diarrhea-like defecation has been defined as a
between day 14 and 7 before starting the study. Moreover, in
defecation frequency ≥3 per day and/or a fecal consistency
the probiotic group one other subject incidentally consumed
≥5 per day, on the Bristol stool form scale, for at least 2
that probiotic during the first 2 wk of the study. Apart from
one subject in the probiotic group taking omeprazole oncedaily on day 45 and 46 of the study, no medication poten-
Statistics
tially affecting the intestinal microbiota was taken during the
The treatment allocation was concealed to all investigators
and volunteers, until the study had been completed and allanalyses had been performed. Bacteriological Culture
The primary outcome of this study was to compare the
During probiotic intake, a significant increase in the mean
changes that occurred in the composition of the intesti-
number of fecal enterococci was found on day 7 (5.8 vs 4.0
nal microbiota during and after amoxycillin intake between
log cfu/g feces, P < 0.02) and on day 14 (6.9 vs 4.3, P <
probiotic- and placebo-treated subjects. Secondary outcomes
0.001) in the probiotic group compared to the placebo group
were the changes that occurred in the metabolic activity of the
(Table 1). Moreover, the mean number of fecal enterococci
intestinal microbiota and changes in defecation score during
within the probiotic group increased significantly during an-
and after amoxycillin between probiotic- and placebo-treated
tibiotic/probiotic intake (day 7) and increased further during
probiotic therapy alone (day 14). A significant decrease in the
Only data from volunteers who completed the study, had
mean number of fecal enterococci was observed 7 wk after
a probiotic/placebo and antibiotic compliance of ≥90%, and
cessation of probiotic intake (P < 0.05) having returned to
delivered all four fecal samples were included in the data
pretreatment level (Table 1). No further differences in either
aerobic or anaerobic bacterial species could be seen between
Statistical evaluation of differences between groups and
the probiotic and the placebo group.
changes within groups (at all time points during the study
However, group-specific differences were observed over
period) was carried out using linear mixed model analysis.
time: within the probiotic group a significant decrease was
In this analysis the fixed effects were day and treatment and
found in total aerobes (day 63 versus day 7) and significant
the random effect was subject. For two-group comparisons
increases were observed over time in total anaerobes (day 14
of independent ordinal and interval values the nonparamet-
versus day 0) and Bacteroides spp. (day 7 and day 14 versus
ric Mann-Whitney U-test was used while the nonparametric
day 0) (P < 0.05). Within the placebo group a significant
Wilcoxon signed-ranked test was used for comparison of re-
increase was found over time in enterococci (day 14 versus
lated ordinal and interval values. All tests were conducted
day 0) and significant decreases were found in lactobacilli
using SPSS version 11.0 (SPSS Inc, Chicago, IL) and a P
(day 7 versus 0) and spore-forming clostridia (day 7 versus
value below 0.05 was considered statistically significant.
day 0 and 63) (P < 0.05) (Table 1). Effect of a Multispecies Probiotic on the Intestinal Microbiota and Bowel Movements Table 1. Numbers of Bacteria Cultured Expressed as Log cfu/g Feces
∗Between group difference P < 0.02. †within group decrease t = 63 versus t = 7, P < 0.05. ‡within group increase t = 7/14 versus t = 0/63 and t = 7 versus t = 14, P < 0.05. §within group increase t = 14 versus t = 0, P < 0.05. ¶within group increase t = 14 versus t = 0, P < 0.05.
within group increase t = 7/14 versus t = 0, P < 0.05.
∗∗within group decrease t = 7 versus t = 0/63, P < 0.05. ††within group decrease t = 7 versus t = 0, P < 0.05.
For all bacterial species studied in both the probiotic and
Metabolic Activity
the placebo group, values on day 63 did not differ significantly
β-Glucosidase activity did not differ significantly between the
probiotic and the placebo group during the total study period
The PFGE profile of 39 out of the 40 enterococci strains,
(Table 2). Within both groups, a decrease in β-glucosidase
isolated from the feces of the healthy volunteers receiving
was observed at day 7 (significant for the placebo group),
probiotic, was similar to that of the orally administrated pro-
which increased again on day 63 (significant for the probiotic
biotic E. faecium W54 strain.
group). In both groups, the β-glucosidase activity returned
Table 2. Metabolic Activity: β-Glucosidase Activity (Expressed as mg/60 min/g Feces); SCFA Concentration (Expressed in mmol/g Feces) and pH β-glucosidase Probiotic∗ Placebo† SCFA Acetic acid Probiotic‡ Placebo§ Probiotic¶ Probiotic Placebo∗∗ Probiotic
∗within group increase t = 63 versus t = 7, P < 0.05 and a tendency to a within group decrease t = 7 versus t = 0, and increase t = 63 versus t = 14, P < 0.06. †within group decrease t = 7 versus t = 0, P < 0.05. ‡within group decrease t = 7/14 versus t = 0, P < 0.05. §within group increase t = 14 versus t = 7, P < 0.05 and within group decrease t = 63 versus t = 14, P < 0.05. ¶within group decrease t = 14/63 versus t = 7.
within group decrease t = 7/14 versus t = 0, P < 0.05 and within group increase t = 63 versus t = 7/14, P < 0.05.
∗∗within group decrease t = 7/63 versus t = 0, P < 0.05. Koning et al. Figure 1. Linear regression of fecal consistency (scored with the Bristol stool form scale) in relation to the protein concentration per gram feces (R = 0.61, P < 0.01). Consistency ranging from 1 (hard lumps) to 7 (watery) according to the Bristol stool form scale. Red dots represent fecal samples from each volunteer collected on day 14.
to baseline values 7 wk after cessation of amoxycillin intake
the probiotic group. This effect was also observed for acetic
acid. Furthermore, an increase in propionic acid concentra-
Endotoxin concentrations (mean ± SEM), expressed as
tions was observed on day 7 in the probiotic group.
log ng/mL fecal water, did not differ significantly between
No significant changes were found in the pH of the fecal
the probiotic and the placebo group on day 0 (2.15 ± 0.07 vs
water between and within both groups, during the total study
2.04 ± 0.08), day 7 (2.30 ± 0.06 vs 2.22 ± 0.07), and day
14 (1.92 ± 0.10 vs 1.89 ± 0.11). However, in both groups, a
At all time points, a negative correlation (P < 0.05) was
small but not significant increase in mean endotoxin concen-
observed between the amount of protein per gram feces and
tration was observed on day 7, whereas 1 wk after cessation
the consistency score (except at t = 63 days). In addition,
of antibiotic intake (day 14) a significant decrease (P < 0.05)
a positive correlation (P < 0.05) was observed between the
in endotoxin concentration was observed compared to day 7.
amount of fecal water per 10 g of feces and the consistency
Clostridium toxin A was detected in the feces of two vol-
score at all time points (an example is shown in Fig. 1).
unteers in the placebo group 1 wk after cessation of antibioticintake (day 14) and in one volunteer in the probiotic group at
Defecation Score
The mean defecation frequency and consistency before an-
No significant differences between the groups were ob-
tibiotic and probiotic intake (day 0, i.e., baseline), during the
served for any of the SCFAs tested (Table 2). However, within
antibiotic/probiotic period (day 1–7), and during the probiotic
both groups changes were found over time. Butyric acid con-
only period (day 8–13) are listed in Table 3. The defecation
centrations significantly decreased in both groups during an-
frequency during the probiotic only period (day 8–13) was
tibiotic intake, but by day 63 had recovered to baseline in
significantly lower (P < 0.05) in the probiotic than in the
Table 3. Mean Daily Fecal Frequency and Consistency Scores Before Antibiotic/Probiotic Intake (Day 0), During Antibiotic/Probiotic Intake (Day 1–7), and During Probiotic Intake Alone (Day 8–13)
∗Consistency ranging from 1 (hard lumps) to 7 (watery) according to the Bristol stool form scale. †Significant difference between probiotic and placebo group (P < 0.05). Effect of a Multispecies Probiotic on the Intestinal Microbiota and Bowel Movements Table 4. Defecation Score Between Day 1 and Day 13 versus the placebo group, apart from a significant increase in
fecal enterococci. Although no other differences were ob-
served between groups, group-specific changes were seenover time. Such changes were also observed for metabolic
activity. Finally, a significantly better defecation score (de-
crease in diarrhea-like bowel movements) was observed in
the probiotic group versus the placebo group.
In AAD, differences in efficacy have been reported for dif-
a consistency† ≥ 5 for at least
ferent bacterial species, bacterial strains, and probiotic mix-
tures (28). The efficacy of multispecies probiotic mixtures is
further supported by the successful use of the multispecies
probiotic VSL#3 in several gastrointestinal disorders (40–
42). In the present study, we used a multispecies probiotic
†Consistency ranging from 1 (hard lumps) to 7 (watery) according to the Bristol stool
containing 10 different probiotic strains selected on the basis
of their in vitro ability to inhibit growth of Clostridium spp. and to survive a low pH (2.5) as well as bile and digestive en-
placebo group. During the total probiotic period (day 1–14),
zymes (pancreatin and pepsin) (data not shown). In addition,
diarrhea-like bowel movements were reported less frequently
their resistance profile against a wide range of antibiotics was
in the probiotic (48%) than in the placebo (79%) group (P <
taken into account to prevent possible transfer of resistance
from the probiotic bacteria to the indigenous microbiota. Fi-nally, the combination of strains chosen was tested to exclude
Side Effects
The placebo and probiotic group were comparable regard-
The composition of the fecal microbiota regarding total
ing the percentage (Fig. 2) and severity of side effects re-
aerobic bacteria, clostridia, and lactobacilli counts before in-
ported: 79% mild-moderate symptoms in the probiotic group
tervention was comparable with previous findings in healthy
versus 90% mild-moderate symptoms in the placebo group.
volunteers using the same culture methods (39, 43). How-
Side effects most frequently reported were nausea, abdomi-
ever, in this study lower mean fecal bacterial concentrations
nal cramps, bloating, and flatulence. Finally, side effects were
were found for total anaerobic bacteria and Bacteroides spp.,
significantly more frequent during the antibiotic/probiotic pe-
riod (day 1–7) than during the probiotic only period (day
One of the possible mechanisms by which probiotics exert
8–13) (P < 0.05), for both the probiotic and placebo group
their effect is by affecting the composition of the intesti-
nal microbiota and preventing the overgrowth of possiblepathogens. Only the fecal microbiota was investigated in thisstudy, even though mucosa-associated bacteria may also be
DISCUSSION
very relevant. Due to interindividual variation and possible
In this placebo-controlled double-blind study, investigating
sampling error, various biopsies ought to have been taken at
the effect of a multispecies probiotic in healthy volunteers
all the different time points of the study. Considering the in-
after amoxycillin intake, no differences in the composition of
vasiveness and the potential risks this was considered not to
the intestinal microbiota were observed in the probiotic group
be ethically acceptable in healthy volunteers.
The consumption of the multispecies probiotic, containing
E. faecium, was associated with a significant increase in the
concentration of fecal enterococci in the probiotic group from
log 4.1 cfu/g to log 5.8 cfu/g on day 7 and to log 6.9 cfu/g
on day 14. This increase disappeared 7 wk after cessation
of probiotic intake, demonstrating that the consumption of
this multispecies probiotic, containing E. faecium, can tran-
siently alter the number of viable enterococci. Considering a
consumption of 1 × 109 cfu E. faecium per day, present in the
multispecies product, and a fecal volume of 100 g per day,
the recovery of around log 7 enterococci per gram feces after
probiotic intake indicates that E. faecium is able to survive
passage through the gastrointestinal tract very well. More-
over, the PFGE patterns of the enterococci isolated from the
Figure 2. Side effects. ∗Significant difference (P < 0.05) in the
fecal samples were similar to the orally administered E. fae-
percentage of volunteers with side effects during day 1–7 versuscium. A study in which a monospecies E. faecium probiotic
(4.5–7.5 × 109 cfu daily) was given to healthy volunteers also
Koning et al.
found a high increase (100-fold) in the total number of ente-
5–20% (52, 53), which can further increase with length of stay
rococci (44). Furthermore, the recovery of viable enterococci
in the placebo group was not affected during amoxycillin in-
A change in the composition of the intestinal micro-
take, indicating that in this study amoxycillin had little effect
biota might affect its metabolic characteristics, such as β-
on the indigenous enterococci population.
glucosidase activity. β-Glucosidase has been implicated in
In contrast to the counts of enterococci, twice daily probi-
carcinogenesis, since it is able to hydrolyse dietary substrates
otic consumption containing 3 × 109 cfu lactobacilli (L. sali-
into carcinogenic compounds (55, 56). A decrease of this ac-
varius, L. plantarum, and L. rhamnosus) did not significantly
tivity is therefore potentially beneficial. It has been demon-
increase the number of fecal lactobacilli. However, a previous
strated that a change in the composition of the intestinal mi-
study by our group, in which 20 healthy volunteers consumed
crobiota or the intake of Lactobacillus spp. can influence
L. plantarum 299v for 4 wk did show a 1000-fold increase
β-glucosidase activity, although the results differ between
in the mean number of fecal lactobacilli (39). An increase in
strains and populations studied (39, 46, 57–61). In the present
the mean fecal number of lactobacilli was also observed by
study β-glucosidase activity decreased in both groups during
others after a 6-month consumption of L. rhamnosus and a 3-
amoxycillin intake and returned to baseline values 7 wk after
wk consumption of L. acidophilus (45, 46). However, during
cessation of amoxycillin. No effect of probiotic intake was
amoxycillin intake a decrease was observed in the total num-
ber of fecal lactobacilli in the placebo group but not in the
The endotoxin concentrations in both groups increased
probiotic group. Comparable results were observed in a study
during amoxycillin intake, though not significantly, and de-
performed by Plummer et al. in which probiotic supplemen-
creased significantly 1 wk after cessation of amoxycillin in-
tation was given during H. pylori eradication therapy (30).
take. This is in accordance with evidence from several studies
These findings indicate that probiotic intake might prevent a
showing that antibiotics increase the bioavailability of endo-
decrease of lactobacilli caused by antibiotic intake.
toxin originating from Gram-negative bacteria (62–65). The
Apart from lactobacilli and the E. faecium, the multispecies
level of intestinal endotoxin, however, does not only correlate
probiotic used also contained bifidobacteria. However, no bi-
with the number of Gram-negative bacteria, which is in line
fidobacteria were cultured, due to insufficient selectivity and
with the fact that no changes were seen in total counts of enter-
sensitivity of media available. In future studies, quantification
obacteriaceae, but can also be associated with the metabolic
of bifidobacteria ought to be performed with molecular-based
activity associated with proliferation (62). The clinical sig-
nificance of antibiotic-induced endotoxin release remains to
Looking at both intervention groups, specific changes dur-
be clarified. There is evidence that endotoxemia may be of
ing and after amoxycillin intake were observed, indicating an
importance in patients with increased gut permeability and
effect of amoxycillin intake on the gut microbiota. These re-
that probiotics show potential in preventing loss of gut bar-
sults are in line with the literature (47–50). The various effects
rier integrity (66, 67). Some studies suggest that a reduction
over time in the probiotic group compared to the placebo
in intestinal endotoxin concentration may be associated with
group suggest that the intake of the multispecies probiotic
decreased endotoxin leakage across the gut wall, and sub-
had an impact on the microbiota during amoxycillin intake,
sequently with the control of endotoxin-related conditions
possibly contributing to the better defecation score. This pro-
(68). In our study, probiotic intake had no effect on intestinal
biotic effect on the microbiota is partly caused by the bacteria
themselves, as is reflected in the increase of enterococci in the
The major SCFAs arising from the bacterial fermenta-
probiotic group. In addition, the increase in, for example, the
tion of nondigestible carbohydrates are acetic acid, propi-
total anaerobic microbiota and the absence of a decrease in
onic acid, and butyric acid. They serve as important energy
the spore-forming clostridia during amoxycillin intake sug-
sources (mostly butyric acid) for colonocytes, are associated
gests that the probiotic bacteria were able to induce a change
with the regulation of water and electrolyte transport, and de-
in the intestinal environment favoring the growth of these
crease colonic pH, thereby inhibiting overgrowth of potential
commensal organisms. The fact that the differences between
pathogens (69). In a study with 31 severe AAD patients dis-
the groups were not significant is probably due to the high
turbances in the intestinal microbiota were observed as was a
reduction of the amounts of all major fecal SCFAs (70). SCFA
Alteration of the colonic microbiota due to antibiotic treat-
concentrations and anaerobic cultural counts also decreased
ment can result in overgrowth of C. difficile in the colon. How-
after systemic ceftriaxone treatment in 10 healthy volunteers
ever, no increase in Clostridium spp. was observed in either
(71). Probiotics, by interacting with the intestinal microbiota
group during or after antibiotic therapy. Moreover, during
and being saccharolytic, can alter SCFA concentrations in the
antibiotic therapy Clostridium toxin A was not detected in
colon. Studies have demonstrated different effects on SCFA
the stool of any of the volunteers. This was to be expected
concentrations after probiotic intake, with some showing no
as the prevalence of C. difficile colonization among healthy
effect (39, 43, 72–74), and others showing either an increase
adults is very low, generally less than 2% (51). The spores
(75, 76) or a decrease in specific SCFA concentrations (61).
of these bacteria are usually acquired from hospitals and
Possible explanations for these inconsistent findings are the
long-term-care facilities where the prevalence ranges from
techniques applied, the populations studied, and the different
Effect of a Multispecies Probiotic on the Intestinal Microbiota and Bowel Movements
probiotic strains used. In the present study, decreased acetic
amoxycillin effect, which differed between the probiotic and
acid and butyric acid concentrations were observed during
the placebo group. Moreover, the intake of a multispecies pro-
antibiotic treatment, only returning to baseline 7 wk after ces-
biotic significantly reduced diarrhea-like bowel movements
sation of antibiotic intake in the probiotic group. Moreover,
in healthy volunteers receiving amoxycillin. Although the
an increased propionic acid concentration was observed in the
changes over time in microbial composition and metabolic
probiotic group at day 7. In contrast, the main fermentation
activity by themselves were small, the sum of potentially ben-
products of the bacteria present in the multispecies probi-
eficial changes may have contributed to the improved defe-
otic are lactate, acetate, and formate (the latter only formed
cation score observed. The present study therefore supports
by bifidobacteria) and do not include propionate. In this re-
the hypothesis that multispecies probiotics could be used in
spect, metabolic cross-feeding is likely to have occurred as
the prevention of AAD, as they affect the composition and
lactate can be converted into butyrate or propionate. Which
function of the intestinal microbiota.
metabolic pathway is utilized depends on the compositionof the microbiota as well as environmental conditions, and
ACKNOWLEDGMENTS
shows high interindividual variation (77, 78). In general, theoverall SCFA concentration seemed to be higher in the probi-
The authors would like to thank Ing. M. Hazen and L. Hoff-
otic group, which could be one of the explanations for the less
man for their technical assistance and B. Winkens, Ph.D.,
diarrhea-like defecation score in this group, due to a better
for his help with the statistical analyses. This research was
water and electrolyte absorption (79). It should be noted that
funded by SenterNovem, an agency of the Dutch Ministry of
only 1–5% of the amount of SCFAs produced is excreted in
Economic Affairs (grant no. TSGE 1041).
the feces and that changes in SCFA concentration can be dueto both changes in production and/or absorption and altered
STUDY HIGHLIGHTS
This study demonstrated that the intake of a multispecies
What Is Current Knowledge
probiotic resulted in a significantly better defecation score
r The disruption of the intestinal microbiota by an-
(decrease in diarrhea-like bowel movements), which is in ac-
tibiotics may result in antibiotic-associated diarrhea
cordance with previous studies showing that probiotics sig-
nificantly reduce the relative risk of developing AAD (26–
r Clinical studies show that probiotics seem efficacious
28). Fecal consistency was estimated by the validated Bristol
stool form scale. The significant correlation of the consis-
r Beneficial probiotic effects differ per probiotic bacte-
tency with both the amount of protein and fecal water per
gram feces supports the validity of this scale.
During antibiotic intake a significant number of side ef-
What Is New Here
fects was reported in both groups, but their numbers did not
r Amoxycillin intake affects both the microbial compo-
differ between the probiotic and the placebo group. These
sition and metabolic aspects of the fecal microbiota.
results suggest that the multispecies probiotic that was able
r The multispecies probiotic Ecologic AAD affects
to decrease diarrhea-like defecation does not reduce other
both the composition as well as the metabolic activ-
gastrointestinal side effects, but also does not result in ad-
ity of the fecal microbiota in healthy volunteers taking
verse events. The clinical relevance of the improved defeca-
tion score has to be further studied in specific patient popula-
r The multispecies probiotic Ecologic AAD causes a
tions who have an increased risk of AAD due to host factors
small but significant reduction in diarrhea-like bowel
(age, immune status), hospitalization status, and exposure to
movements in healthy volunteers taking amoxycillin.
The compliance rates for both antibiotic and probi-
otic/placebo intake were high in this study, although we read-ily admit that this was self-reported. We also acknowledge
Reprint requests and correspondence: C.J.M. Koning, Division of
that many gastrointestinal bacteria remain uncultured and that
Gastroenterology-Hepatology, University Hospital Maastricht, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands.
molecular-based techniques would allow a more complete
Received December 15, 2006; accepted August 5, 2007.
assessment of microbial diversity. However, culture providesinformation on quantitative alterations in viable counts of
REFERENCES
specific groups of bacteria, which are also important for themetabolic activity of the intestinal microbiota.
1. Bartlett JG. Antibiotic-associated diarrhea. Clin Infect Dis
In conclusion, comparing the specially developed multi-
2. McFarland LV. Epidemiology, risk factors and treatments for
antibiotic-associated diarrhea. Dig Dis 1998;16:292–307.
ences were observed in bacterial counts nor in metabolic
3. McFarland LV. Epidemiology of infectious and iatrogenic
activity, apart from an increase in enterococci. However,
nosocomial diarrhea in a cohort of general medicine pa-
changes over time were present in both groups indicating an
tients. Am J Infect Control 1995;23:295–305. Koning et al.
4. Gustafsson A, Berstad A, Lund-Tonnesen S, et al. The effect
23. Tankanow RM, Ross MB, Ertel IJ, et al. A double-blind,
of faecal enema on five microflora-associated characteris-
placebo-controlled study of the efficacy of Lactinex in
tics in patients with antibiotic-associated diarrhoea. Scand
the prophylaxis of amoxicillin-induced diarrhea. DICP
5. Kelly CP, Pothoulakis C, LaMont JT. Clostridium difficile
24. Thomas MR, Litin SC, Osmon DR, et al. Lack of ef-
colitis. N Engl J Med 1994;330:257–62.
fect of Lactobacillus GG on antibiotic-associated diarrhea:
6. Aronsson B, Mollby R, Nord CE. Diagnosis and epidemi-
A randomized, placebo-controlled trial. Mayo Clin Proc
ology of Clostridium difficile enterocolitis in Sweden. J An-
timicrob Chemother 1984;14(Suppl D):85–95.
25. Lewis SJ, Potts LF, Barry RE. The lack of therapeutic effect
7. Hogenauer C, Hammer HF, Krejs GJ, et al. Mechanisms and
of Saccharomyces boulardii in the prevention of antibiotic-
management of antibiotic-associated diarrhea. Clin Infect
related diarrhoea in elderly patients. J Infect 1998;36:171–4.
26. Cremonini F, Di Caro S, Nista EC, et al. Meta-analysis: The
8. Beaugerie L, Petit JC. Microbial-gut interactions in health
effect of probiotic administration on antibiotic-associated
and disease. Antibiotic-associated diarrhoea. Best Pract Res
diarrhoea. Aliment Pharmacol Ther 2002;16:1461–7.
Clin Gastroenterol 2004;18:337–52.
27. D’Souza AL, Rajkumar C, Cooke J, et al. Probiotics in pre-
9. Van Vlem B, Vanholder R, De Paepe P, et al. Immunomod-
vention of antibiotic associated diarrhoea: Meta-analysis.
ulating effects of antibiotics: Literature review. Infection
28. McFarland LV. Meta-analysis of probiotics for the pre-
10. Levy SB. The 2000 Garrod lecture. Factors impacting on the
vention of antibiotic associated diarrhea and the treat-
problem of antibiotic resistance. J Antimicrob Chemother
ment of Clostridium difficile disease. Am J Gastroenterol
11. De Giorgio R, Barbara G, Stanghellini V, et al. Diagnosis
29. Orrhage K, Sjostedt S, Nord CE. Effect of supplements with
and therapy of irritable bowel syndrome. Aliment Pharmacol
lactic acid bacteria and oligofructose on the intestinal mi-
croflora during administration of cefpodoxime proxetil. J
12. Madden JA, Plummer SF, Tang J, et al. Effect of probiotics
Antimicrob Chemother 2000;46:603–12.
on preventing disruption of the intestinal microflora follow-
30. Plummer SF, Garaiova I, Sarvotham T, et al. Effects of
ing antibiotic therapy: A double-blind, placebo-controlled
probiotics on the composition of the intestinal micro-
pilot study. Int Immunopharmacol 2005;5:1091–7.
biota following antibiotic therapy. Int J Antimicrob Agents
13. McAuliffe OE, Klaenhammer TR. Genomic perspec-
tives on probioitcs and the gastrointestinal microflora. In:
31. Timmerman HM, Koning CJ, Mulder L, et al. Monos-
Tannock GW, ed. Probiotics and prebiotics, where are
train, multistrain and multispecies probiotics–A compar-
we going? Norfolk: Caister Academic Press, 2002:263–
ison of functionality and efficacy. Int J Food Microbiol
14. Gotz V, Romankiewicz JA, Moss J, et al. Prophylaxis against
32. Ouwehand AC, Isolauri E, Kirjavainen PV, et al. The mucus
ampicillin-associated diarrhea with a lactobacillus prepara-
binding of Bifidobacterium lactis Bb12 is enhanced in the
tion. Am J Hosp Pharm 1979;36:754–7.
presence of Lactobacillus GG and Lact. delbrueckii subsp.
15. Surawicz CM, Elmer GW, Speelman P, et al. Prevention of
bulgaricus. Lett Appl Microbiol 2000;30:10–3.
antibiotic-associated diarrhea by Saccharomyces boulardii:
33. Gomes AM, Malcata FX, Klaver FA. Growth enhancement
A prospective study. Gastroenterology 1989;96:981–8.
of Bifidobacterium lactis Bo and Lactobacillus acidophilus
16. Wunderlich PF, Braun L, Fumagalli I, et al. Double-blind
Ki by milk hydrolyzates. J Dairy Sci 1998;81:2817–25.
report on the efficacy of lactic acid-producing Enterococ-
34. O’Donnell LJ, Virjee J, Heaton KW. Detection of pseudo-
cus SF68 in the prevention of antibiotic-associated diar-
diarrhoea by simple clinical assessment of intestinal transit
rhoea and in the treatment of acute diarrhoea. J Int Med
35. Hartemink R, Domenech VR, Rombouts FM. LAMVAB–a
17. Siitonen S, Vapaatalo H, Salminen S, et al. Effect of Lacto-
new selective medium for the isolation of lactobacilli from
bacillus GG yoghurt in prevention of antibiotic associated
faeces. J Microbiol Methods 1997;29:77–84.
36. Summanen P, Baron EJ, Citron DM, et al. Wadsworth anaer-
18. McFarland LV, Surawicz CM, Greenberg RN, et al. Preven-
obic bacteriology manual, 5th Ed. Belmont, CA: Star Pub-
tion of beta-lactam-associated diarrhea by Saccharomycesboulardii compared with placebo. Am J Gastroenterol
37. Tenover FC, Arbeit RD, Goering RV, et al. Interpreting chro-
mosomal DNA restriction patterns produced by pulsed-field
19. Vanderhoof JA, Whitney DB, Antonson DL, et al. Lacto-
gel electrophoresis: Criteria for bacterial strain typing. J Clin
bacillus GG in the prevention of antibiotic-associated diar-
rhea in children. J Pediatr 1999;135:564–8.
38. van den Braak N, van Belkum A, van Keulen M, et al.
20. Armuzzi A, Cremonini F, Bartolozzi F, et al. The effect
Molecular characterization of vancomycin-resistant entero-
of oral administration of Lactobacillus GG on antibiotic-
cocci from hospitalized patients and poultry products in The
associated gastrointestinal side-effects during Helicobac-
Netherlands. J Clin Microbiol 1998;36:1927–32. ter pylori eradication therapy. Aliment Pharmacol Ther
39. Goossens D, Jonkers D, Russel M, et al. The effect of
Lactobacillus plantarum 299v on the bacterial composition
21. Arvola T, Laiho K, Torkkeli S, et al. Prophylactic Lacto-
and metabolic activity in faeces of healthy volunteers: A
bacillus GG reduces antibiotic-associated diarrhea in chil-
placebo-controlled study on the onset and duration of ef-
dren with respiratory infections: A randomized study. Pedi-
fects. Aliment Pharmacol Ther 2003;18:495–505.
40. Bibiloni R, Fedorak RN, Tannock GW, et al. VSL#3
22. Adam J, Barret A, Barret-Bellet C. Esais cliniques controles
probiotic-mixture induces remission in patients with active
en doubles insu de l’ultra-leur lyophyilisee. Etude multi-
ulcerative colitis. Am J Gastroenterol 2005;100:1539–46.
centrique par 25 medicines de 38 cas. Gaz Med Francaise
41. Kim HJ, Camilleri M, McKinzie S, et al. A randomized
controlled trial of a probiotic, VSL#3, on gut transit and
Effect of a Multispecies Probiotic on the Intestinal Microbiota and Bowel Movements
symptoms in diarrhoea-predominant irritable bowel syn-
61. Spanhaak S, Havenaar R, Schaafsma G. The effect of
drome. Aliment Pharmacol Ther 2003;17:895–904.
consumption of milk fermented by Lactobacillus casei
42. Mimura T, Rizzello F, Helwig U, et al. Once daily high
strain Shirota on the intestinal microflora and immune
dose probiotic therapy (VSL#3) for maintaining remission
parameters in humans. Eur J Clin Nutr 1998;52:899–
in recurrent or refractory pouchitis. Gut 2004;53:108–14.
43. Goossens D, Jonkers D, Russel M, et al. Survival of the
62. Goris H, de Boer F, Van Der Waaij D. Kinetics of endotoxin
probiotic, L. plantarum 299v and its effects on the faecal
release by Gram-negative bacteria in the intestinal tract of
bacterial flora, with and without gastric acid inhibition. Dig
mice during oral administration of bacitracin and during in
vitro growth. Scand J Infect Dis 1988;20:213–9.
44. Lund B, Edlund C, Barkholt L, et al. Impact on human in-
63. Holzheimer RG. Antibiotic induced endotoxin release
testinal microflora of an Enterococcus faecium probiotic and
and clinical sepsis: A review. J Chemother 2001;13:159–
vancomycin. Scand J Infect Dis 2000;32:627–32.
45. Link-Amster H, Rochat F, Saudan KY, et al. Modulation of
64. Hurley JC. Endotoxemia: Methods of detection and clinical
a specific humoral immune response and changes in intesti-
correlates. Clin Microbiol Rev 1995;8:268–92.
nal flora mediated through fermented milk intake. FEMS
65. Hurley JC. Antibiotic-induced release of endotoxin. A ther-
Immunol Med Microbiol 1994;10:55–63.
apeutic paradox. Drug Saf 1995;12:183–95.
46. Tannock GW, Munro K, Harmsen HJ, et al. Analysis of the
66. Llopis M, Antolin M, Guarner F, et al. Mucosal coloni-
fecal microflora of human subjects consuming a probiotic
sation with Lactobacillus casei mitigates barrier injury in-
product containing Lactobacillus rhamnosus DR20. Appl
duced by exposure to trinitronbenzene sulphonic acid. Gut
Environ Microbiol 2000;66:2578–88.
47. Brismar B, Edlund C, Nord CE. Impact of cefpodoxime
67. Madsen K, Cornish A, Soper P, et al. Probiotic bacteria en-
proxetil and amoxicillin on the normal oral and intestinal
hance murine and human intestinal epithelial barrier func-
microflora. Eur J Clin Microbiol Infect Dis 1993;12:714–9.
tion. Gastroenterology 2001;121:580–91.
48. Christensson B, Nilsson-Ehle I, Ljungberg B, et al. A ran-
68. Van Saene JJM, Stoutenbeek CC, Van Saene HK, et al.
domized multi-centre trial to compare the influence of ce-
Reduction of the inestinal endotoxin pool by three differ-
faclor and amoxycillin on the colonization resistance of the
ent SDD regimens in human volunteers. J Endotoxin Res
digestive tract in patients with lower respiratory tract infec-
69. Cook SI, Sellin JH. Review article: Short chain fatty acids in
49. Gipponi M, Sciutto C, Accornero L, et al. Assessing modi-
health and disease. Aliment Pharmacol Ther 1998;12:499–
fications of the intestinal bacterial flora in patients on long-
term oral treatment with bacampicillin or amoxicillin: A
70. Gustafsson A, Lund-Tonnesen S, Berstad A, et al. Faecal
random study. Chemioterapia 1985;4:214–7.
short-chain fatty acids in patients with antibiotic-associated
50. Leigh DA. Pharmacology and toxicological studies with
diarrhoea, before and after faecal enema treatment. Scand J
amoxycillin, talampicillin and ampicillin and a clinical trial
of parenteral amoxycillin in serious hospital infections.
71. Meijer-Severs GJ, Van Santen E, Meijer BC. Short-chain
fatty acid and organic acid concentrations in feces of healthy
51. Poutanen SM, Simor AE. Clostridium difficile-associated
human volunteers and their correlations with anaerobe cul-
diarrhea in adults. CMAJ 2004;171:51–8.
tural counts during systemic ceftriaxone administration.
52. McFarland LV, Mulligan ME, Kwok RY, et al. Nosocomial
Scand J Gastroenterol 1990;25:698–704.
acquisition of Clostridium difficile infection. N Engl J Med
72. Fukuda M, Kanauchi O, Araki Y, et al. Prebiotic treatment
of experimental colitis with germinated barley foodstuff: A
53. Rivera EV, Woods S. Prevalence of asymptomatic Clostrid-
comparison with probiotic or antibiotic treatment. Int J Mol
ium difficile colonization in a nursing home population: A
cross-sectional study. J Gend Specif Med 2003;6:27–30.
73. Hove H, Nordgaard-Andersen I, Mortensen PB. Effect of
54. Johnson S, Gerding DN, Olson MM, et al. Prospective, con-
lactic acid bacteria on the intestinal production of lactate
trolled study of vinyl glove use to interrupt Clostridium dif-
and short-chain fatty acids, and the absorption of lactose. ficile nosocomial transmission. Am J Med 1990;88:137–40.
55. Goldin BR. In situ bacterial metabolism and colon muta-
74. Rochet V, Rigottier-Gois L, Sutren M, et al. Effects of orally
gens. Annu Rev Microbiol 1986;40:367–93.
administered Lactobacillus casei DN-114 001 on the com-
56. Rowland I. Metabolic interaction in the gut. In: Fuller R,
position or activities of the dominant faecal microbiota in
ed. Probiotics: The scientific basis. London: Chapman and
healthy humans. Br J Nutr 2006;95:421–9.
75. Johansson ML, Nobaek S, Berggren A, et al. Survival of
57. Cole CB, Fuller R, Mallet AK, et al. The influence of the
Lactobacillus plantarum DSM 9843 (299v), and effect on
host on expression of intestinal microbial enzyme activi-
the short-chain fatty acid content of faeces after ingestion of
ties involved in metabolism of foreign compounds. J Appl
a rose-hip drink with fermented oats. Int J Food Microbiol
58. Gadelle D, Raibaud P, Sacquet E. beta-Glucuronidase activi-
76. Schneider SM, Girard-Pipau F, Filippi J, et al. Effects of
ties of intestinal bacteria determined both in vitro and in vivo
Saccharomyces boulardii on fecal short-chain fatty acids
in gnotobiotic rats. Appl Environ Microbiol 1985;49:682–5.
and microflora in patients on long-term total enteral nutri-
59. Goldin BR, Swenson L, Dwyer J, et al. Effect of diet and
tion. World J Gastroenterol 2005;11:6165–9. Lactobacillus acidophilus supplements on human fecal bac-
77. Bourriaud C, Robins RJ, Martin L, et al. Lactate is mainly
terial enzymes. J Natl Cancer Inst 1980;64:255–61.
fermented to butyrate by human intestinal microfloras
60. Marteau P, Pochart P, Flourie B, et al. Effect of chronic in-
but inter-individual variation is evident. J Appl Microbiol
gestion of a fermented dairy product containing Lactobacil-lus acidophilus and Bifidobacterium bifidum on metabolic
78. Louis P, Scott KP, Duncan SH, et al. Understanding the
activities of the colonic flora in humans. Am J Clin Nutr
effects of diet on bacterial metabolism in the large intestine.
J Appl Microbiol 2007;102:1197–208. Koning et al.
79. Wong JM, de Souza R, Kendall CW, et al. Colonic health:
M.A.E. Jonkers were responsible for the analyses of the study.
Fermentation and short chain fatty acids. J Clin Gastroen-
All authors participated in the design and data interpretation
of the study and contributed significantly to the various drafts
80. Macfarlane GT, Macfarlane S. Human colonic micro-
biota: Ecology, physiology and metabolic potential of in-
testinal bacteria. Scand J Gastroenterol Suppl 1997;222:
Financial support: This study was sponsored by Senter-
Novem, an agency of the Dutch Ministry of Economic Affairs(grant no. TSGE 1041). CONFLICT OF INTEREST Potential competing interests: Linda Mulder and Frans M. Guarantor of the article: R.W. Stockbr¨ugger, M.D., Ph.D.
Rombouts are employees of Winclove Bio industries B.V. Specific author contributions: Catherina J.M. Koning was
Catherina J.M. Koning is supported, in part, by a grant from
the principal investigator. Catherina J.M. Koning and Daisy
SenterNovem and, in part, by Winclove Bio Industries B.V.
Koning et al.