Share this article on:

Quantitative Analysis of Intestinal Bacterial Populations From Term Infants Fed Formula Supplemented With Fructo-oligosaccharides

Xia, Qing*; Williams, Timberly; Hustead, Deborah; Price, Pamela; Morrison, Mark*; Yu, Zhongtang*

Journal of Pediatric Gastroenterology and Nutrition: September 2012 - Volume 55 - Issue 3 - p 314–320
doi: 10.1097/MPG.0b013e3182523254
Hepatology and Nutrition

Objectives: Previous studies of infant formulas supplemented with oligosaccharides reported mixed results regarding the impact on intestinal microbial populations. The objective of this study was to examine the effect of supplementation of an infant formula with fructo-oligosaccharides (FOS) on select groups of intestinal bacteria in term infants.

Methods: Four groups of infants were enrolled and fed human milk, a commercially available milk-based infant formula, or infant formula supplemented with 2.0 or 3.0 g/L FOS. Dietary intake, stool, and tolerance events were recorded. Fresh stool samples were collected approximately 27 days after feeding the diets (approximately 32 days after birth). Total bacteria, Bacteroides (as commensal bacteria), Bifidobacterium and Lactobacillus, and Clostridium difficile and Escherichia coli were quantified using respective specific real-time PCR assays.

Results: The formula feeding groups did not differ in stool consistency and stool frequency or frequency of spit-up or vomit during the entire study. The formula-fed infants tended to have more total bacteria in their stool samples than the human milk–fed infants. The formula-fed infants harbored a greater abundance of C difficile and E coli than the human milk–fed infants, but had a similar abundance of Bacteroides, Bifidobacterium, and Lactobacillus. The FOS supplementation at either dose did not significantly increase the bifidobacterial or lactobacilli populations, or decrease the populations of C difficile, E coli, or Bacteroides.

Conclusions: The milk-based formula used in this study supported bifidobacterial and lactobacilli populations comparable with the human milk group; however, this formula did not suppress E coli or C difficile as effectively as human milk.

*Department of Animal Sciences, Ohio State University

Abbott Nutrition, Abbott Laboratories, Columbus, OH.

Address correspondence and reprint requests to Zhongtang Yu, Department of Animal Sciences, Ohio State University, Columbus, OH 43210 (e-mail:

Received 27 March, 2011

Accepted 25 January, 2012

Drs Xia and Williams participated equally in the study.

The research was funded by Abbott Nutrition.

The authors report no conflicts of interest.

Microbiota development in the newborn intestinal tract is an area of renewed research because such microbiotas are now recognized to play important role(s) in the maturation of intestinal and systemic immunity and human health (1,2). Relative to infant formula preparations, human milk (HM) has been found in some, but not all, of the published studies to promote colonization and persistence of bifidobacteria and lactobacilli in infant intestinal tract, which are widely considered to be “beneficial” bacteria, while suppressing undesirable bacterial groups such as Escherichia coli, enterococci, and clostridia (3–6). Breast-fed infants are reported to have lower incidences of allergies, diarrhea, and respiratory and gastrointestinal infections than formula-fed peers (7–10). Although all of the factors contributing to the health-promoting effects of HM remain to be identified and characterized, the types and amounts of oligosaccharides present in HM are believed to be important contributing factors that have a bifidogenic effect, and specifically, promote bifidobacterial growth in infant intestines (8,11,12). A number of studies have attempted to compare the intestinal microbiota between breast-fed and formula-fed infants (5,13–15), and to assess the bifidogenic/prebiotic effects of oligosaccharides, especially fructo-oligosaccharides (FOS) and galacto-oligosaccharides (GOS), when added to infant formulas (16–19). To date, the conclusions drawn from these studies are mixed, however, in terms of clearly establishing the differences in intestinal microbiotas between breast-fed and formula-fed infants and determining the magnitude of the bifidogenic effects of oligosaccharides supplementation of infant formulas.

In most of these previous studies, bifidobacteria and other bacterial groups were enumerated by either cultivation-based methods (16,20,21) or fluorescence in situ hybridization (15,17); however, because these methods rely on selective plating or microscopic enumeration, their accuracy and precision can be compromised by the choice of selective growth medium (22) and/or the variability associated with microscopic counting. Enumeration of specific groups of bacteria by real-time polymerase chain reaction (PCR) assays overcomes these limitations (23), and as such, can offer better sensitivity, accuracy, and precision. In recent studies, specific real-time PCR assays are increasingly used for more accurate and precise quantification of select bacterial genera and species in infant feces (24–26).

There are a variety of infant formulas available on the market and these formulas can differ significantly in nutrient composition. The objective of the present study was to examine the microbiota among 4 groups of term infants receiving HM, a milk-based formula, or the milk-based formula supplemented with 2 different concentrations (2.0 vs 3.0 g/L) of FOS. Specific real-time PCR assays were used to quantify the abundance of total bacteria, beneficial bacteria (Bifidobacterium and Lactobacillus), commensal bacteria (Bacteroides), and potentially detrimental bacteria (E coli and Clostridium difficile). The impact of the FOS supplementation on infant stool consistency and frequency was also examined.

Back to Top | Article Outline


Study Design and Sample Collection

Our clinical trial was a randomized, controlled, and multicenter study. An independent ethics committee/institutional review board reviewed and approved the study protocol and consent forms. Informed consent was obtained from the legally authorized representative of the subjects before their participation in the study. A total of 101 healthy term infants with a birth weight of >2490 g and 0 to 6 days of age were enrolled in the clinical trial consisting of 4 types of feeding. These infants were enrolled in 3 metropolitan areas: Tampa, FL, Cleveland, OH, and Pittsburgh, PA. The control formula (CF) was a cow's-milk–based formula without addition of FOS. The supplemented formulas were the CFs supplemented with either 2.0 g (2.4 g, analyzed) FOS/L (FF-L) or 3.0 g (3.4 g, analyzed) FOS/L (FF-H). Formulas were clinically labeled with an α-numeric code. The study was doubled-blinded at the study sites for the formula-fed subjects. Infants who had been exclusively HM fed were enrolled in the concurrent HM reference group. All of the infants were fed the designated feeding for approximately 4 weeks.

Parents were instructed to keep daily diaries on dietary intake, stool and tolerance events, and medication use for the entire feeding period. The daily diaries were used to develop mean rank stool consistency scores, average daily number of stools, percentage of stools in each consistency category, average formula intakes, and percentage of feedings with spit-up/vomit associated with feedings. Stool consistency values were developed from records in which parents were provided a 5-point scale to describe infant stools as: watery (runny, mostly liquid) = 1; loose/mushy (spread over diaper, mixed with water) = 2; soft (spread over diaper, pasty) = 3; formed (has some shape in diaper, yet moist) = 4; hard (well shaped, dry pellets) = 5. Parents also were asked to record any medication use in the diaries. Diarrhea was not expected to occur in the study, and a specific definition of diarrhea was therefore not provided to parents. If diarrhea did occur, either as perceived by the parent or diagnosed by the study physicians, it was recorded as an adverse event.

Infants were evaluated at enrollment (study day 1), and after 2 and 4 weeks (study visits 2 and 3, respectively) on the assigned study formula. At visits 2 and 3, parents of infants in the formula groups completed questionnaires regarding their satisfaction with the formula, and all of the parents completed questionnaires regarding feeding and stool patterns. Infant body weight was recorded at each study visit. Adverse events were collected throughout the study. Study infants were excluded if they received antibiotics, or if the mother of a HM-fed infant received antibiotics, during the study.

On day 28 (±3 days) of the designated feeding, a fresh fecal sample was collected. Stool samples were collected from the diaper at home and immediately frozen by the infants’ parents, and then delivered to nearby clinical sites (Tampa, FL, Cleveland, OH, and Pittsburgh, PA) within 24 hours, where the stool samples were immediately stored at −20°C before delivery to the laboratory on dry ice for analyses. If the samples were not immediately processed in the laboratory, they were stored at −80°C until DNA extraction, which was completed no later than 2 weeks after sample delivery. All of the stool samples were analyzed in a random order before identification of the treatment groups at the end of all of the analyses.

Back to Top | Article Outline

DNA Extraction

The stool samples were thawed on ice and each of them homogenized with a sterile spatula. Total community DNA was extracted from 0.25 to 0.5 g each of wet stool samples as described previously (27). After visual assessment of DNA quality by agarose gel (0.8%) electrophoresis, the resultant community DNA was quantified fluorospectrometrically using a Quant-it kit (Invitrogen, Carlsbad, CA) on an Mx3000P real-time PCR system (Stratagene, La Jolla, CA).

Back to Top | Article Outline

Quantification of Abundance of Bacterial Groups By Specific Real-time PCR

The primers used in real-time PCR assays were listed in Table 1 (28–34). Because no specific primers targeting the rrs gene of C difficile was published at the time of the present study, a pair of specific primers was designed for this species based on an alignment of all of the C difficile rrs gene sequences then available in the Ribosomal Database Project database ( The primer sequences were compared with database sequences using BLASTn search (GenBank and RDP) and Probe Match (RDP) to ensure match with only the rrs sequences of C difficile. The specificity of this new primer set was confirmed by cloning and sequencing of the PCR products generated using this primer set from one of the infant stool samples (data not shown).



Except for the real-time PCR assays of C difficile and E coli, 1 sample-derived standard was prepared for each real-time PCR assay as described previously (24,35). Briefly, each specific primer set was used to generate a mixture of PCR products from a pooled DNA sample containing an equal amount of DNA from each of the samples to be analyzed. This mixture of PCR products that presumably represented the sequence diversity of the targeted sequences in the samples was confirmed by agarose gel electrophoresis, purified by a Qiagen PCR Product Purification kit (Invitrogen), and quantified by fluorospectrometry (24,35). The genomic DNA of C difficile ATCC 9689D-5 and E coli TOP10 (a derivative of strain K-12) was used to generate the real-time PCR standard for these species by endpoint PCR and universal PCR primers 27f/1525r (Table 1). For each real-time PCR standard, copy number concentration was calculated, based on the length of the PCR product and the mass concentration. The standards were stored at −80°C, and serial dilutions (1–107 copies per microliter) were made before each real-time PCR assay. All of the real-time PCR assays were performed using an Mx3000P real-time PCR system (Stratagene) in 25 μL reaction volumes in triplicate. The real-time PCR master mix contained 1× PCR buffer (20 mmol/L Tris-HCl [pH 8.4] and 50 mmol/L KCl), 200 μmol/L dNTP, 250 nmol/L each primer, 1.75 mmol/L MgCl2, 670 ng/μL bovine serum albumin, 0.133× of SYBR Green I (Invitrogen), 30 nM reference dye ROX (Stratagene), and 1.25 U Platinum Taq DNA polymerase (Invitrogen), which allows hot-start PCR. Total bacteria, Bifidobacterium and Lactobacillus, were quantified using respective specific real-time PCR assays as described previously (24). The abundance of Bacteroides, E coli, and C difficile was quantified similarly. The thermoprofiles of the real-time PCR assays for Bacteroides, E coli, and C difficile were as follows: initial denaturation at 95°C for 4 minutes, followed by 45 cycles of denaturation at 94°C for 30 seconds, annealing for 30 seconds at the respective annealing temperature (shown in Table 1), elongation at 72°C for 40 seconds, and 86°C for 18 seconds. After these 45 cycles, the PCR products were held at 95°C for 1 minute, cooled down to 55°C for 30 seconds, and then slowly ramped up to 95°C, where they were held for 30 seconds to terminate the reactions. Fluorescent signals were acquired twice per cycle: at 72°C and 86°C (endpoint), as well as during the ramping from 55°C to 95°C (all point).

Back to Top | Article Outline

Statistical Analysis and Sample Size Determination

It was anticipated that 16 subjects completing per study formula group would be sufficient to provide meaningful data relative to fecal flora. Therefore, assuming a dropout rate of 20% and 4 candidate study-feeding groups, enrollment of approximately 80 total subjects was anticipated. For each stool sample, data obtained from each real-time PCR assay were converted to average estimate of rrs gene copies per gram of wet stool sample (35). Relative abundance of each bacterial group was calculated from each stool sample as its rrs gene copies per million copies of total bacterial rrs genes. Log10 transformation was applied to all of the absolute and relative abundance values before analyses to improve model assumptions of normality. Likewise, the arcsine square root transformation was used for the percentage of watery stools. All of the continuous data were analyzed with analysis of variance (for data averaged during all 28 days) and repeated measures analysis of variance (for data averaged in 2-week blocks) using the PROC MIXED SAS procedure (SAS Institute, Cary, NC). Least squares means (LSM), standard errors of the mean (SEM), and P values were also calculated. Pairwise comparisons of the LSM were examined with a Tukey-Kramer adjustment made to the P values. Predominant stool consistency and dropout rate were analyzed using the PROC GENMOD SAS procedure. Statistical significance was claimed at P < 0.05.

Back to Top | Article Outline


Demographics and Tolerance

A total of 101 infants were enrolled in the study. Four subjects never received study product and were excluded from the intent-to-treat (ITT) group. There were no significant differences among the 4 feeding groups in sex or birth weight at enrollment for the infants in the ITT group. The age (days) of enrollment (mean ± SEM) was similar for the 3 formula-fed groups (2.5 ± 0.2 days for CF, 2.2 ± 0.3 days for FF-L, 2.3 ± 0.4 days for FF-H), whereas the HM-fed group was younger (1.6 ± 0.2). The average number of days (mean ± SEM) of the designated feeding was 26.0 ± 1.2 days for CF, 22.7 ± 1.90 days for FF-L, 27.3 ± 0.5 days for FF-H, and 27.30 ± 0.3 days for HM. In the 4 feeding groups, 5 infants (CF, 1; FF-L, 3; FF-H, 1; HM, 0) discontinued their designated feeding due to intolerance reported by their parents; 2 infants discontinued the designated feeding at investigator's request due to an adverse event (CF, 0; FF-L, 2; FF-H, 0; HM, 0); and the rest of the dropout infants did not complete the study due to other reasons not related to feeding tolerance. Of the 101 infants enrolled (CF, 26; FF-L, 26; FF-H, 27; HM, 22), 4 infants never received study feedings so were not included in the ITT group. There were no significant differences between the 4 feeding groups for total number of infants who completed the study and who exited the study.

An ITT dataset (CF, 24; FF-L, 25; FF-H, 26; HM, 22) was generated from the daily diaries of the participating parents. This ITT dataset was used to compare the tolerance among the 3 formula-fed groups. Mean intake in the FF-H group was marginally greater than intake by infants in the FF-L group during days 15 to 28 of the study (P = 0.052). During the course of the feeding (days 1–28), mean stool rank consistency tended (P = 0.08) to be lower (ie, softer) in the FF-H versus the control group. There were no differences among treatment groups in the predominant stool consistency and average daily stool number during the entire study. Similarly, there was no significant difference in the frequency of feedings with spit-up or vomit during the entire study.

Back to Top | Article Outline

Abundance of Bacterial Groups and Total Bacteria

A total of 65 subjects (CF, 14; FF-L, 14; FF-H, 20; HM, 17) who completed the study on their assigned study feeding had a valid fecal sample for analysis. The abundance of total bacteria in the HM group appeared to be slightly lower than that in the 3 formula-fed groups (Fig. 1F); however, there were no significant differences among all of the feeding groups. The LSM values of the total bacteria approximated 10 logs of rrs copies per gram sample, which is in the range typical of infant stool samples (36). The range of differences in total bacteria between the HM group and the 3 formula-fed groups was relatively small, with only a 0.47 to 0.65 log difference.



Bifidobacterium and Lactobacillus varied considerably in absolute abundance among the infants within each of the 4 feeding groups, especially within the FF-L group (Fig. 1A and B). The absolute abundances, LSM (±SEM), of bifidobacteria within the HM, CF, FF-L, and FF-H group were 8.49 (±0.35), 9.29 (±0.39), 8.75 (±0.39), and 9.38 (±0.33) logs of rrs copies per gram sample, respectively, exhibiting no significant difference among the 4 feeding groups. Lactobacillus was found to be much less abundant than Bifidobacterium (Fig. 1B), with the LSM ± SEM values for the HM, CF, FF-L, and FF-H groups being 6.77 (±0.55), 7.79 (±0.61), 6.79 (±0.60), and 7.28 (±0.51) logs of rrs copies per gram sample, respectively. Similarly as bifidobacterial abundance, lactobacilli abundance did not differ significantly among the four feeding groups. On a relative basis, the FF-L group tended to have less bifidobacteria than the other 3 groups, but the differences were not significant because of the considerable variations among the individual infants within each feeding group (Fig. 2A). The relative abundance of lactobacilli quantified in the HM, CF, FF-L, and FF-H groups was much lower than that of bifidobacterial (Fig. 2B) and varied among the 4 feeding groups. Again, none of these differences were found to be statistically significant.



The abundance of Bacteroides ranked second to that of Bifidobacterium in 3 of the 4 feeding groups (Fig. 1C). No significant difference in Bacteroides abundance among feeding groups was observed, although the HM group had a higher average abundance than the 3 formula-fed groups. C difficile was found present in all of the infant stool samples and varied in abundance substantially among infants within each feeding group (Fig. 1D), with the LSM (±SEM) values of rrs copies per gram sample for the HM, CF, FF-L, and FF-H groups being 4.16 (±0.43), 4.96 (±0.47), 6.00 (±0.47), and 5.68 (±0.40) logs, respectively. The HM group had less C difficile than the 3 formula-fed groups, but only the difference with the FF-L group reached statistical significance. Among the 3 formula-fed groups, the CF group had lower C difficile than the FF-L and FF-H groups (1.04 and 0.72 logs lower, respectively), but the difference lacked statistical significance. The relative abundance of C difficile present in the FF-L group and the FF-H group was also greater than that present in the HM and CF groups. These differences were not statistically significant owing to the considerable variations within each group (Fig. 2D).

All but 2 infant stool samples (HM, 1; FF-H, 1) were found to contain E coli. The abundance of E coli in the HM group was lower than that of the 3 formula-fed groups, and the 2 FOS-supplemented formula-fed groups (FF-L and FF-H) had a lower abundance of E coli than the CF group (Fig. 1E). Large degrees of variations were observed among different infants within every feeding group. Due to these variations, none of the differences observed among the feeding groups reached statistical significance. On a relative basis, however, the CF group tended to have more E coli than the other 3 feeding groups. Large variations were also observed among infants within every feeding group (Fig. 2E).

Back to Top | Article Outline


The milk-based infant formula used in the present study supported a bifidobacterial population size comparable with that sustained by HM, and the addition of FOS to this formula at either level (2.0 or 3.0 g/L) did not significantly increase the abundance of bifidobacteria. A number of previous studies (13,14,16,37,38) support these findings, although several other studies observed a greater population of bifidobacteria in breast-fed infant stools (39,40). In an in vitro study, Dubey and Mistry (41) did not find any significant changes in maximal counts or generation time for all 4 species of Bifidobacterium tested (B bifidum, B breve, B infantis, B longum) when 5.0 g/L of FOS was added to a soy-based or a casein hydrolysate–based formula. In another in vitro study by the same researchers (42), both a soy- and a milk-based formula supported a similar, but higher, growth rate for B bifidum, B breve, B infantis, and B longum than a casein hydrolysate–based formula. As such, the inconsistent bifidogenic effect of different formulas and supplementations reported in various studies is probably attributable, at least partially, to differences in the formula used. From an ecological perspective, differences in the intestinal microbiota in individual infants probably also contribute to the “inconsistent” bifidogenic effect observed on the same formula or supplementation. This hypothesis is supported by the inverse relation between the initial bifidobacterial abundance and increase in population sizes observed in in vitro cultures with oligosaccharides serving as sole substrates (33,34). This may help explain why supplementation of FOS to the formula in the present study, which resulted in a bifidobacterial population comparable with that of the HM-fed infants, did not further increase bifidobacterial abundance. It is worth pointing out that the infant formula used in the present study was also supplemented with nucleotides. Because nucleotide addition to infant formulas can have an effect on the gut microbiota (3,15), the nucleotides added to the milk-based formula may also have affected the bifidobacterial abundance noted in the present study. No conspicuous differences in the abundance of lactobacilli were found among the 4 feeding groups. This is consistent with previous studies in which supplementation of infant formula with FOS alone at 1.5 and 3.0 g /L or with both FOS and GOS (16,36,37), or polydextrose, GOS, lactulose, and combination thereof (38) did not significantly stimulate intestinal lactobacilli in the fecal samples of term infants.

The absolute abundance of bifidobacteria and lactobacilli as determined in the present study is in the typical range reported for infant fecal samples (14,16,37,45). Our relative abundance values, however, are lower than those reported by some studies, but comparable with other studies in which predominance by bifidobacteria was only found in a small fraction of infants (25,45). Considerable variations among infants were also observed by these researchers. This could be attributable to many factors including the handling of the stool samples, the sensitivity and specificity of the quantitative methods used, and differences in the intestinal microbiotas of the infants recruited in different geographic regions. For example, compared with real-time PCR, fluorescence in situ hybridization considerably overestimates relative abundance of bifidobacteria (13,14,16,37,38). Future extensive studies may help better elucidate the differences among feeding groups and among individual infants of the same feeding. In the present study, Bifidobacterium and Bacteroides, both of which account for up to 90% of the total bacterial population in infant intestines, were included in the evaluation. Additional bacterial groups of the infants, together with intestinal bacteria of their parents and siblings, also should be included in comprehensive evaluations of prebiotics in future studies.

In the present study, the Bacteroides abundance was examined and compared among the 4 feeding groups. A slightly higher abundance of Bacteroides was observed in the HM group than in the 3 formula-fed groups. This finding is consistent with other studies (45,46). In assessing the bifidogenic effect of FOS and GOS, Penders et al (37) noted that formula supplementation with both oligosaccharides failed to further increase bifidobacterial abundance compared with control that also had high bifidobacterial abundance; however, the supplementation led to slight increase in Bacteroides abundance. In another study (46), formula supplementation with FOS resulted in high abundance of bifidobacteria, but Bacteroides also remained high. Thus, a high abundance of Bifidobacterium does not necessarily correspond to a low abundance of Bacteroides in formula-fed infants regardless of prebiotics supplementation. That is, the inverse relation between Bifidobacterium and Bacteroides abundance typically observed in breast-fed infants may not have duplicated in formula-fed infants even with oligosaccharides supplementation. This is not surprising given the ability of some Bacteroides to use FOS and GOS (47) and the stimulation of Bacteroides in batch cultures with FOS or GOS added as the sole substrate (43,44).

From a health perspective, it is desirable to reduce the populations of C difficile and potentially harmful E coli in the intestine of infants. The abundance of C difficile detected in all of the groups was in the range reported by other groups (13,16,48), but the E coli populations observed in the present study were 2 to 3 logs lower than those reported in those studies. A possible explanation for the lower E coli abundance may be related to differences in the socioeconomic status of the study subjects and/or the feedings consumed in the present study. Consistent with previous findings (13,16), the breast-fed infants had less C difficile and E coli than the formula-fed infants. Our results are also supported by other studies (16,37,49), which also showed that feeding of FOS-supplemented infant formula did not reduce clostridial or E coli abundance to the level found in the feces of breast-fed infants. In a separate study, supplementation of a formula with both FOS and GOS significantly increased the bifidobacterial abundance but did not significantly reduce the population of C difficile (46). The effect of feeding on E coli was not assessed in that study. Nonetheless, a high bifidobacterial population may not necessarily correspond to a low population of C difficile. The fecal samples of formula-fed infants typically have a higher pH (∼6.5) (50) than that of breast-fed infants (51). In an in vitro study, the supernatant of bifidobacterial species inhibited the growth of C difficile, but only at pH <5.0 (52). High fecal pH may be a reason why formula-fed infants have abundance of bifidobacteria and lactobacilli comparable to breast-fed infants, but trend toward more C difficile than the breast-fed infants.

Collectively, the milk-based formula used in the present study supported bifidobacterial and lactobacilli populations comparable with that sustained by HM; however, this formula did not suppress E coli or C difficile as effectively as HM. Supplementation of this formula with FOS at 2.0 or 3.0 g/L did not further increase the population of bifidobacteria or lactobacilli or decrease the population of E coli or C difficile. As demonstrated in the present study and other previous studies (5,13,14,36), continuous improvement in infant formula is closing the gap between formula and HM with respect to bifidobacterial and lactobacilli abundance. Future improvement of infant formula should be directed to reduce the abundance of potentially harmful bacteria including E coli and C difficile. Furthermore, a remarkable degree of variation exists in bifidobacterial and lactobacilli populations among infants. Different prebiotics, or mixture thereof, may be needed for infants carrying different “types” of intestinal microbiota to achieve the significant, desired prebiotics effects as that observed in breast-fed infants.

Back to Top | Article Outline


1. Rautava S, Isolauri E. The development of gut immune responses and gut microbiota: effects of probiotics in prevention and treatment of allergic disease. Curr Issues Intest Microbiol 2002; 3:15–22.
2. Caicedo RA, Schanler RJ, Li N, et al. The developing intestinal ecosystem: implications for the neonate. Pediatr Res 2005; 58:625–628.
3. Mountzouris KC, McCartney AL, Gibson GR. Intestinal microflora of human infants and current trends for its nutritional modulation. Br J Nutr 2002; 87:405–420.
4. Fanaro S, Chierici R, Guenini P, et al. Intestinal microflora in early infancy: composition and development. Acta Paediatr 2003; 92:S48–S55.
5. Hascoet JM, Hubert C, Rochat F, et al. Effect of formula composition on the development of infant gut microbiota. J Pediatr Gastroenterol Nutr 2011; 52:756–762.
6. Friedman NJ, Zeiger RS. The role of breast-feeding in the development of allergies and asthma. J Allergy Clin Immunol 2005; 115:1238–1248.
7. Cebra JJ. Influences of microbiota on intestinal immune system development. Am J Clin Nutr 1999; 69:1046S–1051S.
8. Boehm G, Stahl B. Oligosaccharides from milk. J Nutr 2007; 137:847S–849S.
9. Chierici R, Fanaro S, Saccomandi D, et al. Advances in the modulation of the microbial ecology of the gut in early infancy. Acta Paediatr 2003; 92:56–63.
10. Rubaltelli FF, Biadaioli R, Pecile P, et al. Intestinal flora in breast- and bottle-fed infants. J Perinat Med 1998; 26:186–191.
11. Zivkovic AM, German JB, Lebrilla CB, et al. Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A 2011;108(Suppl 1):4653–8.
12. Gueimonde M, Laitinen K, Salminen S, et al. Breast milk: a source of bifidobacteria for infant gut development and maturation? Neonatology 2007; 92:64–66.
13. Penders J, Vink C, Driessen C, et al. Quantification of Bifidobacterium spp., Escherichia coli and Clostridium difficile in faecal samples of breast-fed and formula-fed infants by real-time PCR. FEMS Microbiol Lett 2005; 243:141–147.
14. Sakata S, Tonooka T, Ishizeki S, et al. Culture-independent analysis of fecal microbiota in infants, with special reference to Bifidobacterium species. FEMS Microbiol Lett 2005; 243:417–423.
15. Bezirtzoglou E, Tsiotsias A, Welling GW. Microbiota profile in feces of breast- and formula-fed newborns by using fluorescence in situ hybridization (FISH). Anaerobe 2011;17:478–82.
16. Euler AR, Mitchell DK, Kline R, et al. Prebiotic effect of fructo-oligosaccharide supplemented term infant formula at two concentrations compared with unsupplemented formula and human milk. J Pediatr Gastroenterol Nutr 2005; 40:157–164.
17. Knol J, Scholtens P, Kafka C, et al. Colon microflora in infants fed formula with galacto- and fructo-oligosaccharides: more like breast-fed infants. J Pediatr Gastroenterol Nutr 2005; 40:36–42.
18. van Hoffen E, Ruiter B, Faber J, et al. A specific mixture of short-chain galacto-oligosaccharides and long-chain fructo-oligosaccharides induces a beneficial immunoglobulin profile in infants at high risk for allergy. Allergy 2009;64:484–7.
19. Ben XM, Li J, Feng ZT, et al. Low level of galacto-oligosaccharide in infant formula stimulates growth of intestinal Bifidobacteria and Lactobacilli. World J Gastroenterol 2008; 14:6564–6568.
20. Boehm G, Lidestri M, Casetta P, et al. Supplementation of a bovine milk formula with an oligosaccharide mixture increases counts of faecal bifidobacteria in preterm infants. Arch Dis Child Fetal Neonatal Ed 2002; 86:F178–181.
21. 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; 34:291–295.
22. Apajalahti JHA, Kettunen A, Nurminen PH, et al. Selective plating underestimates abundance and shows differential recovery of bifidobacterial species from human feces. Appl Environ Microbiol 2003; 69:5731–5735.
23. Kim D. Real-time quantitative PCR. Exp Mol Med 2001; 33:101–109.
24. Anderson K, Yu Z, Chen J, et al. Analyses of Bifidobacterium, Lactobacillus, and total bacterial populations in healthy volunteers consuming calcium gluconate by denaturing gradient gel electrophoresis and real-time PCR. Int J Probiotics Prebiotics 2008; 3:31–36.
25. Palmer C, Bik EM, Digiulio DB, et al. Development of the human infant intestinal microbiota. PLoS Biol 2007; 5:e177.
26. Haarman M, Knol J. Quantitative real-time PCR analysis of fecal Lactobacillus species in infants receiving a prebiotic infant formula. Appl Environ Microbiol 2006; 72:2359–2365.
27. Yu Z, Morrison M. Improved extraction of PCR-quality community DNA from digesta and fecal samples. Biotechniques 2004; 36:808–812.
28. Lane DJ. Stackebrandt E, Goodfellow MD. 16S/23S rRNA sequencing. Nucleic Acid Techniques in Bacterial Systematics. New York:John Wiley & Sons; 1991. 115–175.
29. Nadkarni MA, Martin FE, Jacques NA, et al. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 2002; 148:257–266.
30. Bartosch S, Fite A, Macfarlane GT, et al. Characterization of bacterial communities in feces from healthy elderly volunteers and hospitalized elderly patients by using real-time PCR and effects of antibiotic treatment on the fecal microbiota. Appl Environ Microbiol 2004; 70:3575–3581.
31. Bernhard AE, Field KG. Identification of nonpoint sources of fecal pollution in coastal waters by using host-specific 16S ribosomal DNA genetic markers from fecal anaerobes. Appl Environ Microbiol 2000; 66:1587–1594.
32. Satokari RM, Vaughan EE, Akkermans ADL, et al. Bifidobacterial diversity in human feces detected by genus-specific PCR and denaturing gradient gel electrophoresis. Appl Environ Microbiol 2001; 67:504–513.
33. Sabat G, Rose P, Hickey WJ, et al. Selective and sensitive method for PCR amplification of Escherichia coli 16S rRNA genes in soil. Appl Environ Microbiol 2000; 66:844–849.
34. Walter J, Hertel C, Tannock GW, et al. Detection of Lactobacillus, Pediococcus, Leuconostoc, and Weissella species in human feces by using group-specific PCR primers and denaturing gradient gel electrophoresis. Appl Environ Microbiol 2001; 67:2578–2585.
35. Yu Z, Michel FC Jr, Hansen G, et al. Development and application of real-time PCR assays for quantification of genes encoding tetracycline resistance. Appl Environ Microbiol 2005; 71:6926–6933.
36. Veereman-Wauters G, Staelens S, Van de Broek H, et al. Physiological and bifidogenic effects of prebiotic supplements in infant formulae. J Pediatr Gastroenterol Nutr 2011; 52:763–771.
37. Penders J, Thijs C, Vink C, et al. Factors influencing the composition of the intestinal microbiota in early infancy. Pediatrics 2006; 118:511–521.
38. Nakamura N, Gaskins HR, Collier CT, et al. Molecular ecological analysis of fecal bacterial populations from term infants fed formula supplemented with selected blends of prebiotics. Appl Environ Microbiol 2009; 75:1121–1128.
39. Harmsen HJM, Wildeboer-Veloo ACM, 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; 30:61–67.
40. Haarman M, Knol J. Quantitative real-time PCR assays to identify and quantify fecal Bifidobacterium species in infants receiving a prebiotic infant formula. Appl Environ Microbiol 2005; 71:2318–2324.
41. Dubey UK, Mistry VV. Effect of bifidogenic factors on growth characteristics of bifidobacteria in infant formulas. J Dairy Sci 1996; 79:1156–1163.
42. Dubey UK, Mistry VV. Growth characteristics of bifidobacteria in infant formulas. J Dairy Sci 1996; 79:1146–1155.
43. Rycroft CE, Jones MR, Gibson GR, et al. A comparative in vitro evaluation of the fermentation properties of prebiotic oligosaccharides. J Appl Microbiol 2001; 91:878–887.
44. Rycroft CE, Jones MR, Gibson GR, et al. Fermentation properties of gentio-oligosaccharides. Lett Appl Microbiol 2001; 32:156–161.
45. Hopkins MJ, Macfarlane GT, Furrie E, et al. Characterisation of intestinal bacteria in infant stools using real-time PCR and northern hybridisation analyses. FEMS Microbiol Ecol 2005; 54:77–85.
46. Rinne MM, Gueimonde M, Kalliomäki M, et al. Similar bifidogenic effects of prebiotic-supplemented partially hydrolyzed infant formula and breastfeeding on infant gut microbiota. FEMS Immunol Med Microbiol 2005; 43:59–65.
47. Djouzi Z, Andrieux C, Pelenc V, et al. Degradation and fermentation of alpha-gluco-oligosaccharides by bacterial strains from human colon: in vitro and in vivo studies in gnotobiotic rats. J Appl Bacteriol 1995; 79:117–127.
48. Tonooka T, Sakata S, Kitahara M, et al. Detection and quantification of four species of the genus Clostridium in infant feces. Microbiol Immunol 2005; 49:987–992.
49. Brunser O, Gotteland M, Cruchet S, et al. Effect of a milk formula with prebiotics on the intestinal microbiota of infants after an antibiotic treatment. Pediatr Res 2006; 59:451–456.
50. Wilson M. Wilson M. The gastrointestinal tract and its indigenous microbiota. Microbial Inhabitants of Humans. Cambridge, UK:Cambridge University Press; 2005. 251–317.
51. Edwards CA, Parrett AM. Intestinal flora during the first months of life: new perspectives. Br J Nutr 2002; 88 (suppl 1):S11–18.
52. Trejo FM, Minnaard J, Perez PF, et al. Inhibition of Clostridium difficile growth and adhesion to enterocytes by Bifidobacterium supernatants. Anaerobe 2006; 12:186–193.

FOS; infant formula; intestinal microbiota; prebiotics; real-time PCR

Copyright 2012 by ESPGHAN and NASPGHAN