Acidic Oligosaccharides from Pectin Hydrolysate as New
Component for Infant Formulae: Effect on Intestinal Flora,
Stool Characteristics, and pH
*S. Fanaro, †J. Jelinek, †B. Stahl, †G. Boehm, ‡R. Kock, and *V. Vigi
*Neonatal Intensive Care Unit, University of Ferrara, Ferrara, Italy; †Numico Research, Friedrichsdorf, Germany;
and ‡Laborzentrum, Pohlheim, Germany
Objectives: To come even closer to the functional composi-
tion of human milk, acidic oligosaccharides (AOS) from
pectin were added to well known neutral prebiotics (galacto-
oligosaccharides (GOS) and long-chain fructo-oligosaccharides
(FOS)). The effect of AOS and GOS/FOS/AOS on intestinal
flora, stool characteristics as well as acceptance and tolerance
Methods: Human milk contains 75% to 85% neutral and
15% to 25% acidic oligosaccharides. In this prospective, ran-
domized, double blind study, a mixture of 80% neutral oligo-
saccharides (from long-chain galacto- and long-chain fructo-
oligosaccharides) with 20% acidic oligosaccharides derived
from pectin hydrolysis was investigated. Forty-six term infants
were fed a standard formula supplemented with either malto-
dextrin as control (n = 15), or with 0.2 g acidic oligosaccharides
(n = 16), or with the latter plus 0.6 g neutral oligosaccharides
(mixture of galacto- and fructo-oligosaccharides; n = 15). Fecal
flora using plating technique and pH were measured. Stool
characteristics and possible side effects (crying, vomiting, and
regurgitation) were recorded.
Results: There was no difference in the bifidobacteria counts
between the control and the group supplemented with acidic
oligosaccharides alone (8.75 6 0.50 vs. 8.58 6 0.94 log colony
forming units [CFU]/g stool). In infants fed the combination of
(9.61 6 0.70 log CFU/g stool; P , 0.01). The same pattern was
observed with lactobacilli. Stool consistency was softest in
infants fed the complete oligosaccharide mixture, but also in
those fed formula supplemented with acidic oligosaccharides
alone, the stool consistency was significantly softer compared
with the control group. Fecal pH increased in the controls,
remained constant in acidic oligosaccharides alone, and de-
creased in the complete mixture of oligosaccharides group.
Conclusion: There was no difference in growth, crying, vomit-
ing, and regurgitation patterns between the groups. In summary,
acidic oligosaccharides from pectin hydrolysate are well tol-
erated as ingredient in infant formulae but do not affect intes-
tinal microecology. JPGN 41:186–190, 2005. Key Words:
characteristics—Term infants. ? 2005 Lippincott Williams &
The intestinal flora of breast fed infants differs sig-
nificantly from that of bottle fed infants (1). Because the
intestinal flora has been considered an important phys-
iologic factor for different functions of the gut and the
development of the immune system (2), there are several
attempts to mimic the intestinal flora of breast fed infants
also in the bottle fed infant. For more than a decade,
living bacteria (probiotics) have been used to influence
the composition of the intestinal flora (3). More recently,
prebiotics have also been used (4).
Prebiotics are defined as substances that survive the
passage through the small intestine and are specifically
fermented by colonic bacteria (5). In the light of this defi-
milk, the oligosaccharides fraction has been identified
as a major factor that influences intestinal flora (6,7).
In human milk 75% to 85% of the oligosaccharides
are neutral, and 15% to 25% are acidic. The structure
of human milk oligosaccharides is very complex, and
they are therefore not suitable for the production of
an infant formula (8). Searching for alternatives, a mix-
ture of long-chain galacto-oligosaccharides (GOS) and
long-chain fructo-oligosaccharides (FOS), both neutral
oligosaccharides, has been identified as an effective
prebiotic ingredient (8,9).
Received March 11, 2004; accepted May 23, 2005.
Address correspondence and reprint requests to Dr. Gu ¨nther Boehm,
Director, Infant Nutrition Research, Numico Research Germany,
Bahnstr. 14-30, 61381 Friedrichsdorf, Germany. (e-mail: guenther.
One of the authors (GB) works for a company that produces infant
formulas containing GOS and FOS. The authors declare that this has
in no way affected the presentation of the results and the overall
preparation of the manuscript.
Journal of Pediatric Gastroenterology and Nutrition
41:186–190 ? August 2005 Lippincott Williams & Wilkins, Philadelphia
A challenging question is whether acidic oligosaccha-
rides are needed to functionally increase the biologic
activity of the GOS/FOS mixture. Structurally, the acidic
oligosaccharides of human milk are characterized by
their content of sialic acid (9). Although some enzymes
such as sialyltransferases can be used for synthesis of
sialic acidcontainingoligosaccharides (10),these synthetic
oligosaccharides are not suitable for dietary products.
Guggenbichler et al. (11) analyzed the active compounds
of carrot soup according to Moro (12), which has been
used for more than 1 century as dietary treatment for
diarrhea. They identified pectin oligosaccharides result-
ing from cooking of the carrots as the active compounds.
Pectin is a common structural component of all higher
plants, including fruits and vegetables (13), and its acidic
oligosaccharides are mainly based on galacturonic acid
the structure of acidic oligosaccharides found in human
milk, which are characterizedby the content of sialic acid
(8). Cooking pectin-containing vegetables induces the
cleavage of the long-chain pectin polymers into oligo-
saccharides (14). A similar degradation of pectin is used
during the process of clarification of fruit and vegetable
juices by enzymatic hydrolysis of pectin by means of
esterases, hydrolases, or lyases, a widespread process in
from enzymatic pectinolysis, whose size depends on the
enzyme used and on the duration of the reaction. Thus,
these substances also appear during standard technolog-
ical processes in baby food production.
In the present explorative study, a mixture of three oli-
gosaccharide components mimicking the ratio of acidic
and neutral oligosaccharides in human milk were tested
with respect to their effect on acceptance and tolerance,
intestinal flora, and fecal pH as well as possible side
PATIENTS AND METHODS
Vaginally born healthy infants born at term at the University
Hospital of Ferrara, Italy, whose mothers were not able to
provide breast milk, were eligible for the study. Only infants
of mothers without antibiotic treatment were included. The
Ethical Committee of the University of Ferrara approved the
study protocol, and written informed consent was obtained for
Infants enrolled were randomly allocated to one of three
formulae. All three were produced according to a standard
recipe for term infant formulae based on intact cow’s milk
protein with a 60/40 whey/casein ratio. The only difference in
the composition was the supplementation with oligosaccharides.
One group received an experimental formula that was supple-
mented with 0.2 g/dL of acidic oligosaccharides, corresponding
to the concentration of acidic oligosaccharides in human milk
(n = 16). Maltodextrin (0.6 g/dL) was added to have a similar
oligosaccharides were derived from citrus pectin by enzymatic
hydrolysis (supplier: Richter Pharma, Wels,Austria). The second
group (n = 15) received an experimental formula supplemented
with 0.2 g/dL acidic oligosaccharides in combination with
0.6 g/dL of the GOS/FOS mixture, as described previously (9),
mimicking the total concentration of oligosaccharides as well
as the ratio between neutral and acidic oligosaccharides in
human milk (8). In the control group (n = 15), the infants were
fed with a standard formula enriched with 0.8 g/dL maltodextrin
to have the same carbohydrate concentration than in the
The infants were first evaluated when formula feeding was
started (measurement day 1) and subsequently after 6 study
weeks (measurement day 2). At both measurement days, stools
were collected. Fecal flora and pH were determined, and stool
characteristics as well as possible side effects such as crying,
regurgitation, or vomiting were recorded. Breast feeding longer
than 1 week or any antibiotic treatment were defined as ex-
For microbiologic analysis, 0.2 g of a fresh fecal sample was
homogenized in a cryoprotective glycerol transport medium
(glycerol 10 mL, oxoid 0.1 g, H2O ad 100 mL) and immediately
frozen at 280?C. The samples were transported on dry ice (15).
The quantitative determination of bifidobacteria was per-
formed on a bifido-medium (Heipha No. 20580e, Heidelberg,
Rogosa-agar (Heipha No. 2068e, Heidelberg, Germany), as de-
scribed previously (16). For determination of Clostridia sp.,
a sulphite-polymix-milk agar, and for the determination of
Bacteroides, a Schaedler medium (Heipha No. 3181e, Heidel-
berg, Germany) were used, respectively. Aerobic germs (E. coli,
Proteus, Klebsiella, Pseudomonas, Enterobacter, Citrobacter,
Enterococci) weresemiquantified on a Urin3G-Agar II (Heipha
No. 03383e, Heidelberg, Germany). Biochemical differentia-
tion for Proteus, Klebsiella, Pseudomonas, Enterobacter, and
Citrobacter was obtained using the VITEK-identification-system
(Biomerieux, Nu ¨rtingen, Germany).
The pH was measured in diluted samples at room tem-
perature using a Handylab pH meter (Schott Glas, Mainz,
Germany) equipped with an Inlab 423 pH electrode (Mettler-
Toledo, Columbus, OH).
Stool characteristics were recorded with respect to consis-
tency (score 1–5; 1 = watery; 2 = soft; 3 = seedy; 4 = formed;
5 = hard) and frequency. Stool consistency and color were
evaluated by the appearance of the fresh sample used for
analysis. The consistency of each stool sample collected at both
measurement days was recorded, and the mean score obtained
for 1 day was used to measure the stool consistency of that day.
In addition, the incidence of crying (score 1–3; 1 = no crying;
2 = crying in connection with feeding; 3 = crying independent
on meals), regurgitation (score 1–3; 1 = 0; 2 = 1–2; 3 = more
than 2 regurgitation episodes per day) and vomiting (score 1–3;
1 = 0; 2 = 1; 3 = more than 1 vomiting episodes per day) was
recorded by the mother.
For all infants, growth parameters were measured at both
measurement days. Body weight was determined using a scale
with an accuracy of 65 g. The crown-heel length was measured
using a special board for newborn infants with an accuracy of
The statistical software StatView 5.0 (SAS Institute Inc.,
Cary, NC) was used. Group comparisons were performed
parametrically by analysis of variance (ANOVA). In case of sig-
nificance (alpha error , 0.05), post hoc single group compari-
sons were conducted with the Fisher protected least significant
ACIDIC OLIGOSACCHARIDES AS NEW COMPONENT FOR INFANT FORMULAE187
J Pediatr Gastroenterol Nutr, Vol. 41, No. 2, August 2005
difference test. Statistical figures are expressed as means 6
standard deviations (m 6 SD).
Five of 51 infants enrolled into the trial did not
complete the study (2 fed with the experimental formula
containing acidic oligosaccharides alone, 2 fed with the
experimental formula containing neutral and acidic oli-
gosaccharides in combination, and 1 fed with the stan-
dard formula). In all five cases, restart of breast feeding
was the reason for discontinuation of the study.
There was no difference in the clinical characteristics
of the infants among the three groups (Table 1).
weight gain as well as length increment were not different
between the groups (Table 1). There was also no differ-
ence in the incidence of crying, regurgitation, and vomit-
ing episodes (data not shown).
In the group fed the formula supplemented with acidic
oligosaccharides only, there was no difference in the
counts of bifidobacteria at the end of the study period
when compared with the group fed the standard formula.
However, in the group fed the formula supplemented
with acidic and neutral oligosaccharides, the counts of
bifidobacteria were significantly higher than in both other
groups (Fig. 1). There was also a significant increase of
the counts of lactobacilli in the group fed the combina-
tion of acidic and neutral oligosaccharides, whereas
acidic oligosaccharides alone did not influence the counts
For all other bacteria (Clostridia, E.coli, Proteus,
Klebsiella, Pseudomonas, Enterobacter, Citrobacter,
Enterococci) no group difference was observed at the
in the group fed the formula supplemented with acidic
oligosaccharides when compared with the control group.
However, in the group fed the mixture of acidic and neu-
tral oligosaccharides, the stool frequency was signifi-
cantly higher. In parallel, the stool consistency scores
were lowest in the group fed the complete oligosacchar-
ides mixture, but also, in the group fed the acidic
oligosaccharides, the stool consistency was significantly
softer in comparison with the group fed the standard
formula (Fig. 2).
In the control group, the pH increased during the study
period (5.53 6 0.66–6.34 6 0.70). In contrast with this
increase, the pH remained nearly constant in the group
fed the formula supplemented with acidic oligosacchar-
ides (5.91 6 0.61–6.06 6 0.88) and evendecreasedin the
group fed the combination of neutral and acidic
oligosaccharides (5.57 6 0.62–5.23 6 0.37) (Fig. 3).
In the present study, acidic oligosaccharides alone had
no effect on the counts of bifidobacteria but influenced
the stool pH as well as stool consistency. The effect of the
acidic oligosaccharides in combination with GOS/FOS
was similar to that already observed for the GOS/FOS
mixture alone. Moro et al. (17) have previously shown
that bifidobacteria increase in newborn infants receiving
FOS/GOS. The missing effect of acidic oligosaccharides
on the counts of bifidobacteria and lactobacilli might
indicate that acidic oligosaccharides have other functions
than being bifidogenic. In fact, for human milk acidic
oligosaccharides, it is well known that they mainly act
as anti-adhesive factor on pathogens (8). In addition, for
the pectin hydrolysate used in this study, Guggenbichler
et al. (11) demonstrated an anti-adhesive effect of the
ever, the present study was not designed to investigate the
anti-adhesive effect on pathogens in the intestinal flora.
It is well possible that the lack of bifidogenic effect by
active oligosaccharides was related to their low concen-
tration. In a study on term infants using the same plating
technique with the prebiotic GOS/FOS mixture, a con-
centration of 0.4 g/100 mL was significantly less effective
than a concentration of 0.8 g/dL (17). Thus, the acidic
oligosaccharides concentration that was adapted to the
composition of human milk (8) was possibly too low to
show a prebiotic effect. However, for safety reasons, we
would not recommend to exceed the concentration of
acidic oligosaccharides beyond that found in human milk.
There was a significant effect on age-related fecal
pH modifications during the study. Because the pH is
strongly influenced by metabolites of the intestinal flora,
TABLE 1. Most relevant clinical data of the infants studied
Study groupStandard formula +Acidic OS+Acidic +neutral OS
Gestational age (wk)
Weight at birth (g)
Length at birth (cm)
Age at study entry (measurement day 1) (days)
Age at end of the study (measurement day 2) (days)
Weight gain during study period (g/d)
Length gain during study period (cm/wk)
39.7 6 1.9
3502 6 471
50.6 6 1.9
55 6 5.3 (41–63)
35.7 6 7.9
0.95 6 0.22
39.5 6 1.7
3433 6 305
50.3 6 1.2
54 6 7.3 (43–65)
33.3 6 7.1
0.91 6 0.29
39.8 6 1.7
3570 6 426
51.7 6 1.5
55 6 4.6 (46–64)
37.6 6 6.7
0.95 6 0.24
188FANARO ET AL.
J Pediatr Gastroenterol Nutr, Vol. 41, No. 2, August 2005
this indirectly indicates that other changes in the in-
testinal flora than an increased number of bifidobacteria
were induced. This would also be in line with the ob-
served changes in stool consistency, which cannot be
exclusively attributed to an osmotic effect because of the
low concentration of the acidic oligosaccharides.
In summary, the present study demonstrates that pectin
hydrolysates used as a source of acidic oligosaccharides
in a concentration typical for acidic oligosaccharides in
human milk were well tolerated as a supplement to an
infantformula.Although therewas nosignificanteffecton
bifidobacteria, both the effects of the acidic oligosacchar-
ides on stool pH and consistency might indirectly indicate
an effect on the intestinal flora. Further studies are needed
to evaluate the biologic effect of pectin hydrolysate as
ingredient for infant formulae.
Acknowledgment: The authors thank Tatjana Hendrich and
Beate Mu ¨ller-Werner for their excellent work on analyses of
1. Harmsen HJ, Wildeboer-Veloo AC, Raangs GC, et al. Analysis of
intestinal flora development in breast fed and formula fed infants by
using molecular identification and detection methods. J Pediatr
Gastroenterol Nutr 2000;30:61–7.
2. Heavey P, Rowland I. The gut microflora of the developing infant:
microbiology and metabolism. Microb Ecol Health Dis 1999;11:
3. Markowitz JE, Bengmark S. Probiotics in health and disease in the
pediatric patient. Pediatr Clin North Am 2002;49:127–41.
4. Boehm G, Jelinek J, Stahl B, et al. Prebiotics in infant formulas.
J Clin Gastroenterol 2004;38:S76–9.
5. Gibson GR, Roberfroid MB. Dietary modulation of the human
colonic microbiota: introducing the concept of prebiotics. J Nutr
FIG. 2. Stool consistency after a 6 week feeding period in infants
fed a standard term formula (n = 15), a formula supplemented
with 0.2 g/dL acidic oligosaccharides alone (n = 16), or a formula
supplemented with 0.2 g/dL acidic oligosaccharides in combina-
tion with 0.6 g/dL neutral oligosaccharides (n = 15).
FIG. 3. Change of fecal pH during a 6 week feeding period in
infants fed a standard term formula (n = 15), a formula supple-
mented with 0.2 g/dL acidic oligosaccharides alone (n = 16), or
a formula with 0.2 d/dL acidic oligosaccharides in combination with
0.6 g/dL neutral oligosaccharides (n = 15).
FIG. 1. Total bifidobacteria counts in feces after a 6 week
feeding period in infants fed a standard term formula (n = 15),
a formula supplemented with 0.2 g/dL acidic oligosaccharides
alone (n = 16), or a formula supplemented with 0.2 g/dL acidic
oligosaccharides in combination with 0.6 g/dL neutral oligosac-
charides (n = 15).
ACIDIC OLIGOSACCHARIDES AS NEW COMPONENT FOR INFANT FORMULAE189
J Pediatr Gastroenterol Nutr, Vol. 41, No. 2, August 2005
6. Gyo ¨rgy P, Norris RF, Rose CS. A variant of Lactobacillus bifidus Download full-text
requiring a special growth factor. Arch Biochem Biophys 1954;48:
7. Kunz C, Rudloff S, Baier W, et al. Oligosaccharides in human milk:
structural, functional and metabolic aspects. Ann RevNutr 2000;20:
8. Boehm G, Stahl B. Oligosaccharides. In: Mattila-Sandholm T, ed.
Functional Dairy Products. Cambridge: Woodhead, 2003; 203–43.
9. Boehm G, Fanaro S, Jelinek J, et al. Prebiotic concept for infant
nutrition. Acta Paediatr 2003;441:64–7.
10. Lee KB, Lee YC. Transfer of modified sialic acids by Trypanosoma
cruzi trans-sialidase for attachment of functional groups to oligo-
saccharide. Anal Biochem 1994;216:358–64.
11. Guggenbichler JP, De Bettignies-Dutz A, Meissner P, et al. Acidic
oligosaccharides from natural sources block adherence of Escher-
icha coli on uroepithelial cells. Pharm Pharmacol Lett 1997;7:
12. Moro E. Karottensuppe bei ern€ ahrungssto ¨rungen der s€ auglinge.
Mu ¨nchen Med Wschr 1908;31:1637–40.
13. Rolin C, Nielsen BU, Glahn PE. Pectin. In: Dumitriu S. ed.
Polysaccharides: Structural Diversity and Functional Versatility.
New York: Marcel Dekker, Inc., 1998;377–431.
14. Ng A, Waldron KW. Effect of cooking and pre-cooking on cell-wall
chemistry in relation to firmness of carrot tissues. J Sci Food
15. Uhlig H. Enzymes in juice and wine making. In: Uhlig H, ed.
Industrial Enzymes and Their Applications. New York: John Wiley
& Sons, Inc., 1998; 313–6.
16. Fanaro S, Vigi V, Chierici R, et al. Fecal flora measurements of
breast fed infants using an integrated transport and culturing
system. Acta Paediatr 2003;92:634–5.
17. Moro G, Minoli I, Mosca M, et al. Dosage related bifidogenic
effects of galacto- and fructo-oligosaccharides in formula fed term
infants. J Pediatr Gastroenterol Nutr 2002;34:291–5.
190 FANARO ET AL.
J Pediatr Gastroenterol Nutr, Vol. 41, No. 2, August 2005