Fecal Microflora in Healthy Infants Born by Different Methods of Delivery: Permanent Changes in Intestinal Flora After Cesarean Delivery : Journal of Pediatric Gastroenterology and Nutrition

Secondary Logo

Journal Logo

Original Articles

Fecal Microflora in Healthy Infants Born by Different Methods of Delivery: Permanent Changes in Intestinal Flora After Cesarean Delivery

Grölund, Minna-Maija*†; Lehtonen, Olli-Pekka; Eerola, Erkki; Kero, Pentti*

Author Information
Journal of Pediatric Gastroenterology & Nutrition 28(1):p 19-25, January 1999.
  • Free


The neonatal period is crucial for intestinal colonization. Born sterile and undergoing colonization within a few days, infants are an open field for colonization by different types of bacteria. Gestational age, type of delivery, and feeding affect the stool flora of young infants (1), but there is little information on which long-term factors influence the bacterial selection process in the gut of young infants 2.

Infants born vaginally apparently acquire their gut flora from maternal vaginal and fecal flora (3), but the environment also contributes. Within maternity wards, nosocomial spread is shown to exist among healthy newborn infants (4,5). For the colonization of infants born by cesarean delivery (CD), the environment is extremely important (6,7). Likewise, if infants are separated from their mothers for long periods after birth, the environment becomes an important source of colonizing bacteria (5).

Gut colonization is delayed in infants born by CD, and intestinal colonization is consequently abnormal for several weeks (1,8). Most of the studies on gut colonization of infants born by CD have extended only to the first month of life (1,8-10). Overall, however, the long-term stability of the gut flora of newborn infants has been studied systematically in only a few studies 2,11. Such information would clearly be valuable, because changes in the primary gut flora have been associated with gastrointestinal disorders and infantile colic 12-15. We studied the stability of the changes in the fecal flora of newborn infants and the association between certain bacteria and gastrointestinal signs. To study two distinctly different groups of neonates with different gut flora, two study groups were formed: vaginally delivered (VD) infants (n = 34) and infants born by CD to mothers who had received antimicrobial prophylaxis (n = 30). The fecal flora was recorded for 6 months, and gastrointestinal signs were registered daily for 2 months.



The study included 64 healthy newborn infants of healthy mothers who delivered at the Department of Obstetrics and Gynecology, University Central Hospital, Turku, Finland. Thirty-four VD infants and 30 infants delivered by elective CD were enrolled between February 1995 and June 1996 after written informed consent had been obtained from their parents. The enrollment rate was two to four children per week, because sampling necessitated that the birth took place on Monday through Thursday. The mothers who delivered by cesarean birth received prophylactically 2 g intravenous ampicillin 2 hours before the operation. None of the mothers had received any antimicrobial agents within the month that preceded the delivery. After delivery, the newborns were admitted randomly to one of the two maternity wards for healthy newborn infants.

Fecal Samples and Bacterial Cultures

Fecal samples were taken when the infants were 3 to 5 days old (taken at the hospital) and 10, 30, 60, and 180 days old (taken at home). The specimens were collected in plastic containers. If not cultured immediately, the samples were stored at 4°C. The specimens obtained at home were taken to the laboratory by the parents. The mean storage time of the specimens was 10.7 hours (range, 0.5-34.5 hours).

An approximate 300-mg portion of the specimen was weighed, diluted, and homogenized in fastidious anaerobe broth (Lab M, Bury, UK). Serial 100-fold dilutions were made in the same broth. Duplicate samples of 10 µl of each dilution were cultured on a variety of nonselective and selective media (Table 1).

Culture media and dilutions

The MacConkey plates were incubated at 35°C in ambient air for 24 hours. All other cultures were incubated in anaerobic jars with gas-generating kits (Anero Gen, AN35; Oxoid, Basingstoke, UK) at 35°C for 48 hours (fastidious anaerobe agar [FAA], Bacteroides bile esculin agar [BBE] and Clostridium perfringers selective agar), or for 72 hours (Rogosa and modified Petuely's agars). The number of colonies from two parallel plates was counted from a dilution yielding 30 to 300 colony-forming units (CFU)/plate, and average was recorded.

The total number of colonies was counted and recorded from MacConkey and FAA agars. The number of all different types of colonies was counted from BBE, Rogosa, and modified Petuely's agars and the number of distinctive black colonies from Clostridium perfringens agar. All different types of colonies from BBE, Rogosa, and modified Petuely's agars and distinctive black colonies from Clostridium perfringers agar were subcultured for aerotolerance testing and were Gram stained.

Anaerobic gram-negative, esculin hydrolysing rods growing on BBE agar were recorded as bacteria of the Bacteroides fragilis group. Anaerobic and aerotolerant nonbranching, grampositive rods with parallel sides from Rogosa agar were recorded as Lactobacillus-like bacteria (LLB). It has been shown that all colonies growing well on Rogosa agar may be considered lactic acid bacteria; some enterococci and pediococci may show reduced growth (16). Anaerobic and aerotolerant gram-positive rods from modified Petuely's agars were recorded as Bifidobacterium-like bacteria (BLB). Modified Petuely's agar is highly selective and efficient for detection of Bifidobacterium from fecal samples. In a previous work by Tanaka and Mutai (17) 175 fecal strains were detected from the modified Petuely's medium; from these 94% were bifidobacteria, 3% eubacteria, and 3% peptostreptococci.

Clostridium perfringens was identified by its distinct colony and Gram-stain morphology, by its anaerobic nature in aerotolerance testing, and by positive results in the reversed CAMP test (18). The bacterial counts were expressed as the log10 of colony-forming units per gram of wet weight of feces.

Clinical Symptoms

The mothers kept a diary for 2 months of their infants' daily bowel habits, abdominal distension, flatulence, normal and colickly crying, use of antimicrobial agents, and feedings. Crying was defined as colicky if it was a distinctive pain cry, and the infant was difficult to console (19). Abdominal distension and flatulence were given a daily score by the following scale: no signs, a few signs, moderate signs, or heavy signs.

Statistical Analysis

The results were analysed statistically by Fisher's exact test (to compare the number of infants colonized at each time point), Mann-Whitney's rank sum test (to compare bacteria counts of colonized infants at 3 days and 6 months of age), and analysis of variance of repeated measurements (to compare the gastrointestinal signs). A commercial software program (Statistica, version 5.0; Stat Soft, Tulsa, OK, U.S.A.) was used for these calculations. P < 0.05 was considered statistically significant.


The study was approved by the joint Committee of Ethics of the Turku University and Turku University Central Hospital.


The study groups differed slightly by gestational age (VD group, 40 weeks; CD group, 39 weeks; p = 0.04) but the birth weights did not differ statistically (VD group, 3577 g; CD group, 3572 g; p = 0.96). Special attention was paid to the method of feeding the infants. The use of formula was recorded carefully by the mothers. The proportion of infants exclusively breast-fed at 2 months of age or partly breast-fed at 6 months of age did not differ between the study groups (p = 0.2 and 0.8, respectively). Eleven infants received antimicrobial agents during follow-up. All of these therapies were administered when the infants were older than 2 months, and in these infants the subsequent 6-mont fecal samples were excluded from the analysis. (Table 2)

Study subjects

Colonization Rates

With the exception of one infant (in the CD group) who was culture-negative at 3 days of age, all infants were colonized with aerobic enteric bacteria in every culture. The colonization rates of BLB and LLB were lower in the CD group 'han in the VD group after birth. The colonization rate of BLB coincided in these groups by 1 month and of LLB by 10 days of age (Figs. 1 and 2). The colonization rate of LLB in the CD group even exceeded that of the VD group in infants 2 and 6 months of age (Fig. 2). The Clostridium perfringens colonization rate was statistically higher in the CD group than in the VD group at 1 month of age (57% vs. 17% p = 0.003; Fig. 3).

FIG. 1:
The percentage of Bifidobacterium-like bacteria (BLB) colonization in infants aged 3, 10, 30, 60, and 180 days born vaginally and by cesarean delivery. *p < 0.001.
FIG. 2:
The percentage of Lactobacillus-like bacteria (LLB) colonization in infants aged 3, 10, 30, 60, and 180 days born vaginally and cesarean delivery.
FIG. 3:
The percentage of Clostridium perfrigens colonization in infants aged 3, 10, 30, 60, and 180 days born vaginally and by cesarean delivery, *p = 0.003.

The colonization rates of bacteria within the Bacteroides fragilis group differed most markedly of all between the study groups. The Bacteroides colonization rate ranged from 52% to 79% in the VD group (Fig. 4). Only one of the infants in the CD group was Bacteroides positive at 3 days of age. After that, Bacteroides was not recovered in any of the samples from the infants in the CD group before the age of 2 months. In infants 6 months of age, the colonization rate was still statistically lower in the CD group than in the VD group (36% vs. 76%; p = 0.009; Fig. 4).

FIG. 4:
The percentage of Bacteroides fragilis group colonization in infants aged 3, 10, 30, 60, and 180 days born vaginally and by cesarean delivery. *p < 0.001, †p = 0.009.

Colonization Levels

The colonization levels of the different bacteria of the colonized infants are shown in Table 3. The total bacterial counts were significantly lower in the CD group in infants 3 days of age (p = 0.005) and in those 6 months of age (p = 0.03). The aerobic enteric bacterial counts did not differ between the groups. Infants in the CD group had a significantly lower level of BLB when they were 3 days old (p = 0.005) but not when they were 6 months old (p = 0.5). There were no differences in the LLB counts or in the Clostridium perfringens counts between the groups. The amounts of Bacteroides fragilis bacteria did not differ between the groups of infants with colonization at 6 months (p = 0.3).

Median and range of fecal bacterial counts of colonized infants at 3, 10, 30, 60, and 180 days of age

Gastrointestinal Signs

Two mothers in both groups did not complete the follow-up sheets. No statistically significant differences were found between the study groups in the scores of abdominal distension, flatulence, or in the amount of colicky crying, Infantile colic, according to the definition of Wessel et al. (20), was detected in three of the VD infants and in none of the CD infants (p = 0.2; Fisher's exact test).


This study shows that the fecal flora of infants born by CD with prophylactic antibiotics administered to the mother is very different from that of infants delivered vaginally. The greatest differences were seen in the bacteria of the Bacteroides fragilis group, which is in agreement with studies in infants born by CD; in earlier studies the duration of follow-up, however, was only 10 to 60 days (1,8,10). In the present study, no permanent colonization with Bacteroides fragilis group bacteria was found in the CD group before the infants were 2 months of age. Still, in infants 6 months of age in the CD group, the colonization rate of the Bacteroides fragilis group was only half that of infants in the VD group (36% and 76%, respectively; p = 0.009). Similarly, Bennet and Nord (8) were unable to detect Bacteroides bacteria 3 to 8 weeks after birth in term and preterm infants born by CD.

We used selective culture media to describe intestinal flora in this study. The counts detected from these selective culture media represent mainly the bacterial genera level but cannot detect the absolute counts of different bacterial genera or species, except those of Clostridium perfringens, which were further identified. Nevertheless, this method can identify differences between the delivery groups and the changes taking place in intestinal flora after birth, which was the main purpose of the current study. Furthermore, the main differences in the fecal flora between the delivery groups were found in Bacteroides fragilis group bacteria that were detected from the BBE media, which is highly selective for the Bacteroides fragilis group (21).

The long-lasting changes seen in the primary gut flora of infants born by CD could be the result of one or both of the two abnormal components of their birth: CD itself or the prophylactic ampicillin and administered to the mother 2 hours before the elective CD. Ampicillin is very poorly protein-bound and crosses the placenta readily: maternal and fetal serum ampicillin levels equilibrate within 1 hour after intravenous administration (22). The decline in BLB could be explained by the intravenous ampicillin that the infants were exposed to before delivery, but fecal LBL should not decline with ampicillin therapy, at least this is not the case in older children (10 months to 12 years). Also, only a minor decline is detected in fecal Bacteroides sp. counts in these older children and the Bacteroides sp. counts return to normal stage within only 3 to 6 days after cessation of the ampicillin therapy (23). Thus, it seems unlikely that the long-term changes recorded in the fecal flora of the infants in the CD group in our study could be explained by the administration of ampicillin before delivery. Rather, previous reports imply that this phenomenon is caused by CD itself (1,8). Still, ampicillin could have more profound effects when administered immediately after the birth because the gut is empty, and the drug can select the primary colonizing bacteria.

The changes that took place in the fecal bacterial flora could not be associated with gastrointestinal signs. Any signs were meticulously recorded daily by the mothers for 2 months. In a previous study, Clostridium perfringens has been associated with an increased incidence of gastrointestinal signs, such as flatulence, distended abdomen, foul-smelling stools, diarrhea, and blood in stools (12). In the present study, the infants born by CD had a higher colonization rate of Clostridium perfringens than the VD group of infants at 1 month (57% and 17%, respectively). Even at this time point, the scores of gastrointestinal signs did not differ between the study groups. The inconsistency of the results between these two studies might be explained by the differences between the study populations. In the previous study most of the infants were treated in the intensive care unit and were preterm infants, whereas in the present study all infants were healthy full-term neonates.

There is great variation among the reports of the predominant bacteria in the fecal flora of breast-fed newborns. Some researchers have found bifidobacteria to predominate and Bacteroides to be scarce (2,24,25), whereas others have found the opposite (26,27). However, the predominant bacterium has been classified differently among the studies. Some investigators define the predominant bacterium as the most frequently occurring bacteria (26) and some as the bacterium with the highest counts in fecal samples (28). In the present study, 19 of the 34 VD infants were exclusively breast-fed for 2 months. Among all the VD children, the most frequent bacteria during the first 2 months of life were BLB (85-97%), although Bacteroides fragilis group bacteria were also common (52-79%). The highest colony counts were encountered for BLB (10.2-10.9 log10 CFU/g wet weight of feces). The Bacteroides colony counts were lower (8.7-9.5 log10 CFU/g). These results agree with those in a recent study from Germany (28).

The present study is the first one to show that the changes in the primary intestinal flora of infants born by CD to mothers who have received antimicrobial prophylaxis last for no less than 6 months after birth, maybe longer. We do not know how long these abnormalities in the fecal flora last on the whole, but the primary quality and quantity of colonization of the gut seems to be critical in the selection process between different genera of bacteria. Some bacteria such as Bacteroides are not as easily accessible to the gut of infants delivered the under sterile conditions of CD. Other, perhaps more aerotolerant, bacteria may take over in the intestine and inhibit the subsequent colonization of the gut by Bacteroides. This phenomenon is the generally known as interbacterial inhibition (29). Apparently, the bacterial predominances are not readily subject to change under normal domestic conditions after the first months of the colonization process.

Normal intestinal flora has immunostimulatory functions, as has been demonstrated in numerous animal studies (30-32). Mucosal IgA plasma cells are especially scarce in germ-free animals (33). Additionally, when probiotic bacteria belonging to the normal intestinal flora have been administered orally to children in association with diarrhea or mucosal vaccination, an increase in antigen-specific and nonspecific IgA and IgM responses has been detected (34,35). Secretory IgA and IgM are the main humoral mediators of mucosal immunity in cooperation with a variety of innate protective mechanisms. Well-functioning mucosal immunity is a prerequisite for health, because the mucosal surfaces are favored as portals of entry by most infectious agents, allergens, and carcinogens (36). Further research is under way to determine whether the delay found in intestinal colonization in these infants born by CD has any effect on the development of the gut-associated immune system.

Acknowledgement: The authors thank Mr. Hans Helenius, M.Sc., Department of Biostatistics, University of Turku, for his help with the statistical analysis; and the nurses of the Department of Obstetrics and Gynecology, and Mrs. Satu Ekblad, Department of Pediatrics, for their invaluable help in performing this study.

Supported in part by The South-West Finnish Fund of Neonatal Research.


1. Long SS, Swenson RM. Development of anaerobic fecal flora in healthy newborn infants. J Pediatr 1977;91:298-301.
2. Stark PL, Lee A. The microbial ecology of the large bowel of breast-fed and formula-fed infants during the first year of life. J Med Microbiol 1982;15:189-203.
3. Tannock GW, Fuller R, Smith SL, Hall MA. Plasmid profiling of members of the family Enterobacteriaceae, lactobacilli, and bifidobacteria to study the transmission of bacteria from mother to infant. J Clin Microbiol 1990;28:1225-8.
4. Delmeé M, Verellen G, Avesani V, Francois G. Clostridium difficile in neonates: Serogrouping and epidemiology. Eur J Pediatr 1988;147:36-40.
5. Murono K, Fujita K, Yoshikawa M, et al. Acquisition of nonmaternal Enterobacteriaceae by infants delivered in hospitals. J Pediatr 1993;122:120-5.
6. Fryklund B, Tullus K, Berglund B, Burman LG. Importance of the environment and the faecal flora of infants, nursing staff and parents as sources of gram-negative bacteria colonizing newborns in three neonatal wards. Infection 1992;20:253-7.
7. Lennox-King SM, O'Farrell SM, Bettelheim KA, Shooter RA. Escherichia coli isolated from babies delivered by caesarean section and their environment. Infection 1976;4:139-45.
8. Bennet R, Nord CE. Development of the faecal anaerobic microflora after caesarean section and treatment with antibiotics in newborn infants. Infection 1987;15:332-6.
9. Hall MA, Cole CB, Smith SL, Fuller R, Rolles CJ. Factors influencing the presence of faecal lactobacilli in early infancy. Arch Dis Child 1990;65:185-8.
10. Neut C, Bezirtzoglou E, Romond C, Beerens H, Delcroix M, Noel AM. Bacterial colonization of the large intestine in newborns delivered by cesarean section. Zentralbl Bakteriol Mikrobiol Hyg A 1987;266:330-7.
11. Ellis-Pegler RB, Crabtree C, Lambert HP. The faecal flora of children in the United Kingdom. J Hyg Lond 1975;75:135-42.
12. Ahtonen P, Lehtonen OP, Kero P, Eerola E, Hartiala K. Clostridium perfringens in stool, intrapartum antibiotics and gastrointestinal signs in a neonatal intensive care unit. Acta Paediatr 1994;83:389-90.
13. Lehtonen L, Korvenranta H, Eerola E. Intestinal microflora in colicky and noncolicky infants: Bacterial cultures and gas-liquid chromatography. J Pediatr Gastroenterol Nutr 1994;19:310-4.
14. Miller JJ, McVeagh P, Fleet GH, Petocz P, Brand JC. Breath hydrogen excretion in infants with colic. Arch Dis Child 1989;64:725-9.
15. Moore DJ, Robb TA, Davidson GP. Breath hydrogen response to milk containing lactose in colicky and noncolicky infants. J Pediatr 1988;113:979-84.
16. The Oxoid Manual. 7th ed. Basingstoke, UK: Unipath Ltd., 1995: 2-186.
17. Tanaka R, Mutai M. Improved medium for selective isolation and enumeration of Bifidobacterium. Appl Environ Microbiol 1980;40:866-9.
18. Buchanan AG. Clinical laboratory evaluation of a reverse CAMP test for presumptive identification of Clostridium perfringens. J Clin Microbiol 1982;16:761-2.
19. Lester BM, Boukydis CFZ, Garcia-Coll CT, Hole WT. Colic for developmentalists. Infant Mental Hlth J 1990;321-33.
20. Wessel MA, Cobb JC, Jackson EB, Harris GSJ, Detweiler AC. Paroxymal fussing in infancy, sometimes called "colic." J Pediatr 1954;14:421-34.
21. Livingston SJ, Kominos SD, Yee RB. New medium for selection and presumptive identification of the Bacteroides fragilis group. J Clin Microbiol 1978;7:448-53.
22. Bray RE, Boe RW, Johnson WL. Transfer of ampicillin into fetus and amniotic fluid from maternal plasma in late pregnancy. Am J Obstet Gynecol 1966;96:938-42.
23. Sakata H, Fujita K, Yoshioka H. The effect of antimicrobial agents on fecal flora of children. Antimicrob Agents Chemother 1986;29:225-9.
24. Roberts AK, Chierici R, Sawatzki G, Hill MJ, Volpato S, Vigi V. Supplementation of an adapted formula with bovine lactoferrin: 1. Effect on the infant faecal flora. Acta Paediatr 1992;81:119-24.
25. Yoshioka H, Iseki K, Fujita K. Development and differences of intestinal flora in the neonatal period in breast-fed and bottle-fed infants. Pediatrics 1983;72:317-21.
26. Lundequist B, Nord CE, Winberg J. The composition of the faecal microflora in breastfed and bottle fed infants from birth to eight weeks. Acta Paediatr Scand 1985;74:45-51.
27. Simhon A, Douglas JR, Drasar BS, Soothill JF. Effect of feeding on infants' faecal flora. Arch Dis Child 1982;57:54-8.
28. Kleessen B, Bunke H, Tovar K, Noack J, Sawatzki G. Influence of two infant formulas and human milk on the development of the faecal flora in newborn infants. Acta Paediatr 1995;84:1347-56.
29. Sprunt K, Redman W. Evidence suggesting importance of role of interbacterial inhibition in maintaining balance of normal flora. Ann Intern Med 1968;68:579-90.
30. Crabbé PA, Bazin H, Eyssen H, Heremans JF. The normal microbial flora as a major stimulus for proliferation of plasma cells synthesizing IgA in the gut. The germ-free intestinal tract. Int Arch Allergy Appl Immunol 1968;34:362-75.
31. Moreau MC, Ducluzeau R, Guy-Grand D, Muller MC. Increase in the population of duodenal immunoglobulin A plasmocytes in axenic mice associated with different living or dead bacterial strains of intestinal origin. Infect Immun 1978;21:532-9.
32. Shroff KE, Meslin K, Cebra JJ. Commensal enteric bacteria engender a self-limiting humoral mucosal immune response while permanently colonizing the gut. Infect Immun 1995;63:3904-13.
33. Crabbé PA, Nash DR, Bazin H, Eyssen H, Heremans JF. Immunohistochemical observations on lymphoid tissues from conventional and germ-free mice. Lab Invest 1970;22:448-57.
34. Kaila M, Isolauri E, Soppi E, Virtanen E, Laine S, Arvilommi H. Enhancement of the circulating antibody secreting cell response in human diarrhea by a human Lactobacillus strain. Pediatr Res 1992;32:141-4.
35. Isolauri E, Joensuu J, Suomalainen H, Luomala M, Vesikari T. Improved immunogenicity of oral D × RRV reassortant rotavirus vaccine by Lactobacillus casei GG. Vaccine 1995;13:310-2.
36. Brandtzaeg P. Molecular and cellular aspects of the secretory immunoglobulin system. APMIS 1995;103:1-19.

In the Next Issue

Original Articles

Sarcosidase Therapy for Congenital Sucrase-Isomaltase Deficiency

William R. Treem, Luanne McAdams, Lesley Stanford, Geraldine Kastoff, Christopher Justinich, and Jeffrey Hyams

Helicobacter pylori Infection in Children with Celiac Disease: Prevalence and Clinicopathologic Features

F. Luzza, M. Mancuso, M. Imeneo, L. Mesuraca, A. Contaldo, L. Giancotti, A. La Vecchia, C. Docimo, L. Pensabene, P. Strisciuglio, F. Pallone, and S. Guandalini

Cisapride in Pediatric Chronic Constipation Management

Issam M. Halabi

Home Parenteral Nutrition in Children: The Polish Experience

Janusz Ksiazyk, Malgorzata Lyszkowska, Jaroslaw Kierkus, Krzysztof Bogucki, Anna Ratyńska, Bozenna Tondys, and Jerzy Socha

Case Report

Congenital Microvillus Atrophy in a Girl with Autosomal Dominant Hypochondroplasia

Peter Heinz-Erian, Heinrich Schmidt, Martine Le Merrer, Alan D. Phillips, Wieland Kiess, and Hans-Beat Hadorn


Bacteroides; Cesarean delivery; Colic; Colonization; Gut; Intestine

© 1999 Lippincott Williams & Wilkins, Inc.