The normal intestinal microflora in newborns is one means of protecting infants against enteric infections. Bacterial colonization of infants begins during parturition as bacteria from the environment compete for nutrients and space in the previously sterile infant gastrointestinal tract (1,2). Breast-feeding affects the sequential colonization of the infant gastrointestinal tract, allowing certain beneficial forms to predominate (3,4). Initially, aerobic and facultatively anaerobic bacteria from the mother’s reproductive and gastrointestinal tracts colonize the infant. As breast-feeding is initiated, anaerobic species soon predominate, especially members of the genera Bifidobacterium and Lactobacillus (3,5). Later, when infants begin to eat solid food, the microflora undergo a final shift to an adult profile with high numbers of anaerobic and relatively low numbers of aerobic and facultatively anaerobic species (3). This general pattern has been described in humans and other animals (6–9). This study is the first to investigate the maturation of intestinal microflora in infant monkeys and its response to prenatal stress.
The prenatal uterine environment has many lasting influences on the developing infant. Prior studies have linked maternal stress during gestation to reduced fetal growth and birth weight (10,11), abnormal development of neuromotor reflexes, emotional lability, and poor cognitive function (12). However, the pervasiveness and duration of these prenatal effects in humans are controversial because of the many different experimental paradigms and outcome measures that have been used to study them. Because rhesus monkeys are similar to humans in fetal development and maturational state at birth, they have been used extensively as models to investigate the influence of prenatal maternal nutrition, alcohol consumption, and stress on infant development. In the monkey, prenatal stress results in delayed neonatal neuromotor development (13), alterations in the newborn’s emotional reactivity to stress (14,15), and changes in brain monoamine levels in older juveniles (16). Maternal disturbance during pregnancy also has an adverse influence on lymphocyteproliferative and cytolytic activity in the infant monkey (17). Even at 2 years of age, prenatally stressed rhesus juveniles have decreased cytokine secretion in response to in vitro lipopolysaccharide stimulation compared with unstressed animals (18).
The intestinal microflora are thought to provide a barrier against enteric pathogens. Although the protective mechanisms are not understood, some intestinal microflora such as Lactobacillus and Bifidobacterium species inhibit pathogen attachment to intestinal cells in vitro and in vivo (19,20). These bacteria also produce antibacterial substances that kill or slow the replication of enteric pathogens (21–24). In addition, changes in the microflora during antibiotic therapy are thought to be responsible for antibiotic-associated diarrheas (2,25). We hypothesized that prenatal stress would decrease the numbers of protective microflora and increase the frequency of infant infection with Shigella flexneri, an opportunistic pathogen endemic in most monkey colonies.
MATERIALS AND METHODS
Twenty-four male and female rhesus monkey infants (Macaca mulatta) were generated from multiparous females that were part of an established breeding colony at the Harlow Center for Biological Psychology (Table 1). The dams had defined pedigrees and known experimental histories and were mated during a 4-day period at the time of ovulation to confirm the date of conception. The gravid females were then housed individually under standardized laboratory conditions until the experimental protocol was begun. Control females were housed with minimal disturbances during both the pre- and postpartum periods. Females assigned to the 6-week experimental stressor (see below) were not disturbed subsequently for the remainder of their pregnancies, nor during the postpartum period. Standard laboratory monkey chow (PMI Nutrition International, Richmond, IN, U.S.A., and Harlan Teklad, Madison, WI, U.S.A.) was given twice daily, with fresh fruit supplements given three times weekly. Water was available ad libitum. All of the females were housed in similar cages (0.8 × 0.8 × 0.8 m). The light:dark schedule was 16:8 with lights on at 0600. All experimental procedures were approved by the Animal Care and Use Committee at the University of Wisconsin.
An acoustical startle was used as the stressor. The pregnant female was moved to a darkened room in a small transport box (0.5 × 0.3 × 0.4 m) daily for 10 minutes between 1430 and 1600 hours. A computer program randomly broadcast three 110-decibel beeps of 1-second duration during the 10-minute stress session. This paradigm was repeated five times per week for 6 weeks. The early stress period lasted from day 50 to 92 of gestation and the late stress period spanned day 105 to 147 of gestation. These periods were chosen because they represent critical periods of nervous system and gastrointestinal tract development (16,26). After the 6-week stress period, the females were left undisturbed in their home cages until the natural birth of their infants. A term pregnancy typically lasts 169 days in the rhesus monkey.
Both groups of dams experienced the same mild disturbances of cage cleaning and routine care. All infants were raised by their biologic mothers in the same facility. However, control animals were housed in a separate room and were undisturbed throughout the study, except for the drawing of four blood samples to match those collected from the stressed mothers. Offspring of the undisturbed pregnancies were then compared with offspring whose mothers had been stressed either in early or late gestation (Table 1).
To determine whether the acoustical startle paradigm had activated the pregnant female’s endocrine physiology, blood samples were collected the week before the acoustic stress period, with three additional samples taken at 2-week intervals during the 6-week stress period. Blood samples were collected between 1500 and 1600 hours on Fridays. Baseline blood samples were collected from the control females at the same stage of gestation and time of day so that all of the females experienced similar handling. Plasma cortisol levels were determined by an iodinated radioimmunoassay (Incstar, Stillwater, MN, U.S.A.).
Enumeration of Microflora
Microflora were enumerated by culturing stool obtained by rectal swab at 2 days of age and at 2, 8, 16, and 24 weeks. Rectal swabs were immersed in 1 mL of prereduced and pre-weighed thioglycollate medium (Remel, Lexington, KY, U.S.A.). After the sample was weighed, it was homogenized in the thioglycollate medium by gently rolling the swab against the side of the test tube until all fecal matter was suspended in the tube. The sample was then serially diluted in prereduced phosphate buffered saline (PBS) to a final dilution of 1:160,000. Aliquots of the dilutions were pour- or spread-plated on selective and differential agars to identify and enumerate different microflora.
Before use, anaerobic agar plates were prereduced for 24 hours at room temperature in anaerobic canisters containing an anaerobic atmosphere of 85% N2, 10% CO2, and 5% H2 generated with a disposable gaspak (Becton Dickinson, Cockeysville, MD, U.S.A.). Total anaerobic bacteria were enumerated by spread-plating on Schaedler agar (Becton Dickinson) with a thin layer of freshly prepared Schaedler agar poured over the plates to prevent swarming of bacterial colonies. Members of the genus Lactobacillus were enumerated with selective LBS agar (Becton Dickinson) and Bifidobacterium spp. were enumerated by spread plating on Bifidobacterium iodoacetate medium 25 (BIM-25). As described previously, BIM-25 agar consists of reinforced clostridial agar (51 g/L), nalidixic acid (0.02 g/L), polymyxin B sulfate (0.0085 g/L), kanamycin sulfate (0.05 g/L), iodoacetic acid (0.025 g/L), and 2,3,5-triphenyltetrazolium chloride (0.025 g/L) (27). After the plates were incubated in an anaerobic atmosphere at 37°C for 3 days, colonies of bacteria growing on the agar were counted, and the number of colony-forming units (CFU) per gram of fecal matter was calculated.
Total aerobic and facultatively anaerobic bacteria were enumerated by pour plating with brain-heart infusion agar (BHI, Becton Dickinson). Eosinmethylene blue agar (EMB, Becton Dickinson) was used to enumerate Gram-negative aerobes and facultative anaerobes via pour-plate analysis. In addition to the intestinal microflora, pathogenic Shigella flexneri, endemic in most monkey colonies, were enumerated by spread-plating on Salmonella-Shigella agar (SS, Remel) and on hektoen enteric agar (HE, Becton Dickinson). After the BHI, EMB, SS, and HE agar plates were incubated at 37°C in an atmosphere of air + 5% CO2 for 24 hours, colonies of bacteria growing on the BHI and EMB agars were counted and the number of CFU/g determined. Suspected colonies of S. flexneri were identified by biochemical reactions in triple sugar iron agar slants (Difco Laboratories, Detroit, MI) and by the agglutination test with Shigella antiserum poly group B (Difco Laboratories).
Mother–infant interactions were observed at 5-minute intervals three times weekly for 8 weeks during the second and third months of life. The duration and frequency of mother and infant behaviors were recorded on a laptop computer between 1400 and 1600 hours. In particular, the number of times mothers pushed their infants away when infants were trying to maintain contact (maternal rejections), and the number of times the mothers retrieved their infants and initiated contact (maternal retrievals), were counted as a measure of infant independence. The time at which the infants first ate solid chow was recorded because ingesting solid food is associated with profound alterations of intestinal microflora (28).
Data were analyzed with one-way and mixed-factor analyses of variance (ANOVA) with prenatal condition (i.e., Control, Early Stress, and Late Stress) as a between factor. For cortisol analyses, blood from some controls was collected early and from others late in gestation to match the two stress periods, but all control infants were considered together for the microflora analyses. Weeks of gestation were treated as a repeated measure in the cortisol analyses. In addition, the age of the infants at the time of microflora enumeration was a repeated measure in the mixed-factor ANOVAs (i.e., 2 days and 2, 8, 16, and 24 weeks). Microflora counts were log(10) transformed before statistical testing and are expressed as the log(10) number of colony-forming units per gram of fecal mass (CFU/g). Post-hoc analyses were performed with protected means comparisons or by using Fisher’s LSD method.
A χ2 analysis was used to determine whether S. flexneri colonization was more prevalent in the prenatally stressed infants. To determine whether mother–infant interactions were good predictors of microflora concentrations in the infant, the relationship between these behaviors at 2 to 3 months and infant microflora at 4 months of age was tested with the Pearson correlation coefficient.
Maternal Cortisol, Gestation Length, and Infant Weight Gain
Cortisol levels were assessed at 2-week intervals in control and stressed females during the period of manipulations. Mean plasma cortisol was elevated in the pregnant dams that were disturbed (F (3, 17) = 5.21, P < 0.01), demonstrating that the acoustical startle paradigm was sufficient to cause a moderate activation of the pituitary-adrenal axis (Fig. 1). Post-hoc testing revealed that females stressed early or late in gestation had significantly higher levels of cortisol than did control animals across the 6-week stress periods (P < .05). Despite elevations in the dam’s cortisol levels, the prenatal stressor did not significantly affect the length of gestation (F (2, 21) = 1.12, P = 0.35, NS) (Table 1). Infant birth weight (F (2, 21) = 0.86, P = 0.45, NS) and weight gain during the first 24 weeks of life (F (2, 21) = 1.88, P = 0.18, NS) also were not different in prenatally stressed infants (Table 1).
Developmental changes in intestinal bacteria occurred during the first 6 months of life in both control and prenatally stressed offspring, as indicated by a significant main effect for infant age in the analyses (F (4, 84) = 10.63, P < 0.0001; F (4, 84) =6.62, P < 0.001, total and Gram-negatives respectively, see Table 2). Fisher’s LSD post-hoc testing revealed that total aerobe and facultative anaerobe counts reached their nadir at 8 weeks (P < .01 in comparison with all other groups) and did not rise again until 16 weeks, when infants began to consume solid foods. However, the levels never exceeded those seen between 2 days and 2 weeks of age. In contrast to this biphasic pattern, Gram-negative aerobes and facultative anaerobes decreased progressively during the breast-feeding period until bacterial counts became significantly lower at 16 and 24 weeks than they had been at 2 days of age (P < .05) (Table 2).
Overall, anaerobic bacteria had the opposite developmental pattern, with relatively low levels on the day after birth and increasing levels through 24 weeks of age. The maturational trend for total anaerobes was not statistically significant (F (4, 84) = 1.74, P = 0.15) (Table 2), but the increases were very pronounced when two specific genera were enumerated. Lactobacilli rose dramatically by age 2 weeks and continued to increase through 16 to 24 weeks (F (4, 84) = 24.79, P < 0.0001) (Fig. 2). Bifidobacteria also increased markedly (F (4, 84) = 4.23, P < 0.01), with counts peaking at 8 weeks (P < .05 in comparison to day 2 levels), after which they declined with the onset of weaning and eating of solid foods (Fig. 3).
The numbers of lactobacilli were significantly different in monkeys from stressed pregnancies, as indicated by a significant main effect for experimental group in the analysis (F (2, 21) = 4.78, P < 0.05) (Fig. 2). Post-hoc tests showed that infants from the undisturbed pregnancies had significantly higher numbers of lactobacilli than did monkeys from either the early or late stress pregnancies (P < .05). This difference was a generalized effect across the first 24 weeks of life because the interaction between the prenatal condition and the age of the animal was not statistically significant (F (8, 84) = 1.81, P = 0.09). Bifidobacteria levels were also affected by stress during pregnancy (F (2, 21) = 3.43, P = 0.05), but here the effect was most evident in infants from late stress pregnancies. Bifidobacteria counts in infants from late stress pregnancies were significantly lower than controls for the first 24 weeks of life (P < .05). However, the early gestational stress did not significantly affect the development of intestinal bifidobacteria (P = .67 v controls). There was a nonsignificant trend for fewer bifidobacteria in infants from late stress pregnancies compared with to infants from early stress pregnancies (P = .07).
Prenatal stress did not affect the total anaerobic bacterial levels (F (2, 21) = 0.56, P = 0.58) or the Gram-negative aerobes and facultative anaerobes (F (2, 21) = 0.33, P = 0.72) (Table 2). There was a trend in both the early and late stress groups to have fewer total aerobes and facultative anaerobes (F (2, 21) = 2.98, P = 0.07).
Although no monkeys were intentionally infected for this study, 43% of the early stress and 13% of the late stress infants shed S. flexneri from their intestines at least once during the 24-week period. In contrast, we did not find S. flexneri in any of the control infants (data not shown). This difference in the prevalence of S. flexneri shedding among the different groups was not statistically significant (χ2 (2) = 5.33, P = 0.07). However, two additional late stress infants were dropped from the study because they required antibiotic treatment for diarrheic symptoms and had stool cultures positive for S. flexneri. Had these two infants been included, the prevalence of S. flexneri in the late stress condition would have been 30%, and the infection level of the prenatally stressed infants would have been significantly higher than that of controls (χ2 (2) = 4.13, P = 0.05). Numbers of lactobacilli were significantly lower in the infant monkeys that became colonized with S. flexneri (mean = 8.3 log(10) CFU/g, SE = 0.1) compared with uncolonized infants (mean = 8.8 log(10) CFU/g, SE = 0.08) (F (1, 24) = 7.37, P < 0.05).
Association Between Microflora Levels and Maternal Behavior
Prenatal stress did not affect the onset of eating solid chow or the amount of time the infants spent eating (F (2, 18) = 0.66, P = 0.53). The progress of the infants toward behavioral independence by 12 weeks of age and their mothers’ response to their attempts to move away from her did relate to the microflora profiles. The ratio of maternal retrievals and rejections in all animals was assessed during the 2nd and 3rd months of life and correlated with total anaerobes (r = .52, P < 0.01), anaerobic lactobacilli (r = .43, P < 0.05), and bifidobacteria concentrations when the infants were 4 months of age (r = .47, P < 0.05) (Fig. 4). This relationship between bacteria and the mother–infant relationship was seen in both control and prenatally stressed monkeys. The total number of retrievals and rejections did not differ as a result of either early or late gestational stress (maternal rejections, (F (2, 21) = 0.69, P = 0.53; retrievals, (F (2, 21) = 1.03, P = 0.37). Bacterial counts did not correlate with infant weight at any time during the first 24 weeks of life (data not shown).
These results support previous studies indicating that even modest disturbances of pregnancy can affect fetal and postpartum infant functions. In humans, the effects of maternal stress during pregnancy are usually most evident as prematurity or intrauterine growth retardation. In many animals, including rodents, ungulates, and primates, infants may be affected even when the newborn is full term and of normal weight (29,30). Previous investigators have focused mainly on behavioral measures, such as the maturity of neuromotor reflexes at birth, or on endocrine measures, but our study indicates that there may be alterations in other systems. In this case, daily disturbance of the gravid female for 25% of the pregnancy significantly affected the establishment of the newborn’s intestinal microflora. Lower levels of the anaerobes normally associated with breast-feeding were found during the first 6 months of life.
The reduction of both lactobacilli and bifidobacteria in infants from late stress pregnancies seems particularly important because these bacteria have been linked to the risk of diarrheal illness, food allergy, and inflammatory bowel diseases in humans (31,32). In the current study, infants born after disturbed pregnancies had a tendency for colonization by pathogenic S. flexneri, which was symptomatic in two infants. These two infants from late stress pregnancies were excluded from all of the other analyses because they required antibiotic treatment. Had these infants been included in the analysis, the incidence of Shigella would have been 30% in infants from late stress and 43% in infants from early stress pregnancies, a prevalence higher than the zero occurrence in control infants from unstressed pregnancies. These infants also had lower Lactobacillus counts during the first 24 weeks of life when compared with uninfected monkeys.
Our experimental stressor reliably raised maternal cortisol above levels normally found during pregnancy, indicating that this prenatal manipulation induced a stress response throughout the 6-week period. Nevertheless, it would be most appropriate to characterize this as a moderate stressor because the maternal cortisol levels were below those found after activating adrenal secretion in the gravid female with adrenocorticotrophic hormone (ACTH) (33) (Fig. 1). It is tempting to speculate that the increased placental transfer of maternal cortisol to the fetus was the mediator of the observed effects on the infants. In fact, it is possible to re-create many influences of prenatal stress by administering steroidal drugs, such as dexamethasone or ACTH to the pregnant dam (17,33, 34). Moreover, injecting pregnant rats with cortisone has been associated with reduced concentrations of total and Gram-negative aerobes and facultative anaerobes, and decreased lactobacilli in the small intestine of pups after birth (35,36). Although there is a prepartum increase in corticosteroid levels in precocial animals that stimulates intestinal maturation (26,37), when very high or sustained, corticosteroids can adversely affect intestinal development by altering gastric acid secretion and the density of villi and crypts on the surface of the small intestine (26,38,39). These changes in intestinal digestive potential or motility could mediate the effects of gestational stress on microflora found in our study (40).
It is possible that stress-induced changes in maternal bacteria or maternal milk might also have contributed to the bacterial changes we observed, but we did not test for this possibility. In addition, it is known that breast milk contains bacteria and that milk constituents facilitate the growth of some bacteria, such as bifidobacteria (41–43). Breast milk contains high concentrations of immune cells (neutrophils, macrophages, colostral corpuscles, and lymphocytes), as well as cytokines (γ-interferon, transforming growth factor β) (44,45) and other immune products such as lactoferrin and immunoglobulin A (IgA) that help to control microflora concentrations throughout the life span. Although no studies have systematically evaluated the effect of gestational stressors on these factors in breast milk, it is known that postpartum stressors can alter secretory IgA levels in breast milk (46). Because the growth of bifidobacteria is highly influenced by the composition of maternal milk (42,43) and because bifidobacteria concentrations were affected in late stress, but not early stress infants (Fig. 3), it is possible that the stress experienced later in pregnancy affected milk production or quality in a way that was no longer evident in the mothers stressed earlier in gestation. Our findings indicate a need to investigate whether pregnancy conditions affect the makeup of maternal milk.
Studies in rodents have demonstrated that the effects of intrauterine stress on the infant are often mediated and extended by long-term changes in the interactions between the mothers and pups (47). For monkey infants, a critical developmental transition begins at 2 months of age with the onset of independent departures off the mother’s ventrum. Our behavioral observations indicate the mother’s retrievals or rejections were highly predictive of bacterial profiles at 4 months of age. The relationship between gut bacteria and the behavioral processes we observed supports a previous study in rhesus monkeys that showed that purposefully disrupting the mother–infant bond by separating infants from dams significantly reduced the number of intestinal lactobacilli in the infants (48).
Our observations on the effects of pregnancy stress contribute to a larger discussion about whether the antecedents of some adult diseases are to be found in fetal life (49). If the intestinal flora in other species are similarly affected by in utero events, this indicates an important and novel pathway requiring additional investigation.
The authors thank Ms. Heather Crispen, Allison Dalal, and Sarah Short for their expert technical assistance.
1. Brook I, Barrett CT, Brinkman CR 3rd, et al. Aerobic and anaerobic bacterial flora of the maternal cervix and newborn gastric fluid and conjunctiva: a prospective study. Pediatrics
2. Tannock GW, Fuller R, Smith SL, et al. Plasmid profiling of members of the family Enterobacteriaceae
, and bifidobacteria
from mother to infant. J Clin Microbiol
3. Cooperstock MS, Zed AJ Intestinal microflora
of infants. In: Hentges D, ed. Human Intestinal Microflora in Health and Disease.
New York: Academic Press; 1983:79–99.
4. Kleesen B, Bunke H, Tovar K, et al. Influence of two infant formulas and human milk on the development of the faecal flora in newborn infants. Acta Paediatr
5. Lejeune C, Bourrillon A, Boussougant Y, et al. Sequential development of the intestinal flora in newborn infants: a quantitative differential analysis. Dev Pharmacol Ther
6. Karney TL, Johnson MC, Ray B. Changes in the lactobacilli
and coliform populations in the intestinal tract of calves from birth to weaning. J Anim Sci
7. Moughan PJ, Birtles MJ, Cranwell PD, et al. The piglet as a model animal for studying aspects of digestion and absorption in milk-fed human infants. In: Simopoulos AP, ed. Nutritional Triggers for Health and in Disease.
Basal: Karger; 1992:40–113.
8. Schaedler RW, Dubos R, Castello R. The development of the bacterial flora in the gastrointestinal tract of mice. J Exp Med
9. Smith HW, Crabb WE. The faecal bacterial flora of the animals and man: its development in the young. J Pathol Bacteriol
10. Field T, Sandberg D, Quetel TA, et al. Effects of ultrasound feedback on pregnancy
anxiety, fetal activity, and neonatal outcome. Obstet Gynecol
11. Lederman RP, Lederman E, Work BA Jr., et al. Maternal psychological and physiological correlates of fetal–newborn health status. Am J Obstet Gynecol
12. Grimm VE, Frieder B. The effects of mild maternal stress
on the behavior of rat pups. Int J Neurosci
13. Schneider ML, Coe CL. Repeated social stress
impairs neuromotor development of the primate infant. J Dev Behav Pediatr
14. Clarke AS, Wittwer DJ, Abbott DH, et al. Long-term effects on prenatal stress
on HPA axis activity in juvenile rhesus monkeys. Dev Psychobiol
15. Clarke AS, Schneider ML. Prenatal stress
has long-term effects on behavioral responses to stress
in juvenile rhesus monkeys. Dev Psychobiol
16. Schneider ML, Clarke AS, Kraemer GW, et al. Prenatal stress
alters brain biogenic amine levels in primates. Dev Psychopathol
17. Coe CL, Lubach GR, Karaszewski JW. Prenatal stress
and immune recognition of self and nonself in the primate neonate. Biol Neonate
18. Coe CL, Kramer M, Kirschbaum C, et al. Prenatal stress
diminishes cytokine-production after an endotoxin challenge and induces glucocorticoid resistance in juvenile rhesus monkeys. J Clin Endocrinol Metab
19. Bernet MF, Brassart D, Neeser JR, et al. Adhesion of human bifidobacterial strains to cultured human intestinal epithelial cells and inhibition of enteropathogen-cell interactions. Appl Environ Microbiol
20. Bernet MF, Brassart D, Neeser JR, et al. Lactobacillus acidophilus
LA1 binds to cultured human intestinal cell lines and inhibits cell attachment and cell invasion by enterovirulent bacteria. Gut
21. Bernet-Camard M, Lievin V, Brassart D, et al. The human Lactobacillus acidophilus
strain LA1 secretes a nonbacteriocin antibacterial substance(s) active in vitro and in vivo. Appl Environ Microbiol
22. Velraeds MMC, Van Der Mei HC, Reid G, et al. Inhibition of initial adhesion of uropathogenic Enterococcus faecalis
by biosurfactants from Lactobacillus
isolates. Appl Environ Microbiol
23. Gibson GR, Wang X. Regulatory effects of bifidobacteria
on the growth of other colonic bacteria. J Appl Bacteriol
24. Rolfe RD, Helebian S, Finegold SM. Bacterial interference between Clostridium difficile
and normal fecal flora. J Infect Dis
25. Arvola T, Laiho K, Torkkeli S, et al. Prophylactic Lactobacillus
GG reduced antibiotic-associated diarrhea in children with respiratory infections: a randomized study. Pediatrics
26. Trahair JF, Robinson PM. The development of the ovine small intestine. Anat Rec
27. Munoa FJ, Pares R. Selective medium for isolation and enumeration of Bifidobacterium
spp. Appl Environ Microbiol
28. Stark PL, Lee A. The microbial ecology of the large bowel of breast-fed and formulae-fed infants during the first year of life. J Med Microbiol
29. Wadhwa PD. In: Friedman HA, ed. Encyclopedia of Mental Health,
vol. 3. San Diego: Academic Press; 1998.
30. Wadhwa PD, Sandman CA, Garite TJ. The neurobiology of stress
in human pregnancy
: implications for prematurity and development of the fetal central nervous system. Prog Brain Res
31. Beale B. Probiotics: their tiny worlds are under scrutiny. Scientist
32. Kirjavainen PV, Arvola T, Salminen E, et al. Aberrant composition of gut microbiota of allergic infants: a target of bifidobacterial therapy at weaning. Gut
33. Coe CL, Lubach GR, Karaszewski JW, et al. Prenatal endocrine activation alters postnatal cellular immunity in infant monkeys. Brain Behav Immun
34. Matthews SG. Antenatal glucocorticoids and programming of the developing CNS. Pediatr Res
35. Wenzl HH, Schimpl G, Feierl G, et al. Effect of prenatal corticosterone on spontaneous bacterial translocation from gastrointestinal tract in neonatal rat. Dig Dis Sci
36. Schiffrin EJ, Carter EA, Walker WA, et al. Influence of prenatal corticosteroids on bacterial colonization in the newborn rat. J Pediatr Gastroenterol Nutr
37. Trahair JF, Sangild PT. Systemic and luminal influences on the perinatal development of the gut. Equine Vet J
38. Sangild PT, Hilsted L, Nexo E, et al. Secretion of acid, gastrin and cobalamin-binding proteins by the fetal pig stomach: developmental regulation by cortisol. Exp Physiol
39. Trahair JF, Perry RA, Silver M, et al. Studies on the maturation of the small intestine of fetal sheep. I. The effects of bilateral adrenalectomy. Q J Exp Physiol
40. Drasar BS. In: Manuel P, Walker-Smith JA, Tomkins A, eds. Infections of the Gastrointestinal Tract
. Edinburgh: Churchill Livingstone; 1986:1–9.
41. Matsumiya Y, Kato N, Watanabe K, et al. Molecular epidemiologic study of vertical transmission of vaginal Lactobacillus species from mothers to newborn infants in Japanese, by arbitrarily primed polymerase chain reaction. J Infect Chemother
42. Liepke C, Aderman K, Raida M, et al. Human milk provides peptides highly stimulating the growth of bifidobacteria
. Eur J Biochem
43. Orrhage K, Nord CE. Factors controlling the bacterial colonization of the intestine in breastfed infants. Acta Paediatr
44. Eglinton BA, Roberton DM, Cummins AG. Phenotype of T-cells, their soluble receptor levels, and cytokine profile of human breast milk. Immunol Cell Biol
45. Saito S, Yoshida M, Ichijo M, et al. Transforming growth factor-beta (TGFB) in human milk. Clin Exp Immunol
46. O’Connor ME, Schmidt W, Carroll-Pankhurst C, et al. Relaxation training and breast milk secretory IgA. Arch Pediatr Adolesc Med
47. Liu D, Dioria J, Tannenbaum B, et al. Maternal care, hippocampal glucocorticoid receptors, and hypothalamic–pituitary–adrenal responses to stress
48. Baily MT, Coe CL. Maternal separation disrupts the integrity of the intestinal microflora
of the infant rhesus monkey
. Devel Psy
49. Saunders PR, Santos J, Hanssen NPM, et al. Physical and psychological stress
in rats enhances colonic epithelial permeability via peripheral CRH. Dig Dis Sci