3.2.1 Development of the intestinal microbiota in the first years of life - role of breast feeding, prebiotics and infant diet
Current data suggest that the intestinal bacterial microbiota undergoes substantial changes in the first year(s) of life and remains relatively stable thereafter, although factors such as acute or chronic intestinal diseases, antibiotic treatment and the composition of the diet influence the composition of the microbiota (Barbut and Petit 2001; Hooper and Gordon, 2001; Nowrouzian et al., 2003; Tannock, 2001 (see Fig. 3.1).
Endogenous mechanisms for control of bacterial growth in the intestine include gastric acid production, secretion of gastric, intestinal and pancreatic proteases such as pepsin, trypsin, chymotrypsin or lysozyme, bile-derived detergents, secretion of intestinal bactericidal peptides and IgA (Eckmann, 2005; Fahlgren et al., 2003; Mowat, 2003; Sarker and Gyr, 1992).
While the intestine in utero is sterile under physiological conditions, a postnatal stepwise colonisation contributes to the development of a normal immune system (Caicedo et al., 2005; Edwards and Parrett, 2002; Fanaro et al., 2003; Guarner and Malagelada, 2003). The maternal vaginal and intestinal microbiota is a source of colonisation of the child (Fanaro et al., 2003). Initially the child's intestine becomes colonised by pioneering enterobacteria. After changing the growing conditions for bacteria due to feeding of breast milk instead of the initial colostrum and due to the metabolic activity of resident pioneer bacteria, Bifidobacteria gain ground. The stool of breast fed infants contains predominately bifidobacteria and less frequently lactobacilli, bacteroides, enterobacteria, enterococci and clostridia (Fanaro et al., 2003). Stepwise colonisation of the gut by defined bacterial strains is important for the development of the intestinal and systemic immune system and mediate protection against nosocomial infections. Bifidobacteria and lactobacilli inhibit
Dynamic interplay between intestinal bacteria, host and environment 'environment'
flora of mother, mode of birth, breast feeding, immune status of mother, diet, antibiotics
Dynamic interplay between intestinal bacteria, host and environment 'environment'
flora of mother, mode of birth, breast feeding, immune status of mother, diet, antibiotics
growth of pathogenic microorganisms through competition for space, production of lactic, acetic and other organic acids as well as H2O2 and antimicrobic peptides (Agostoni et al., 2004; Cross, 2002; Howie, 2002; Howie et al., 1990; Vicente et al., 2003).
The specific microbiota of breast-fed children has been explained by the composition of proteins, the complexity of oligosaccharides and by numerous humoral and cellular mediators (Agramonte-Hevia et al., 2002) within the human milk. There are several factors that influence the composition of the bacterial microbiota and contribute to infection control. Beneficial effects of breast milk are partially mediated via different proteins such as immunoglobulins, kappa-casein, lysozyme, lactoferrin, haptocorrin, alpha-lactalbumin, and lactoperoxidase. Those proteins have anti-microbial activity and support the immune defense of breast-fed infants against pathogenic bacteria and viruses (Lonnerdal, 2003). Further proteins such as insulin-like growth factor, epidermal growth factor or lactoferrin are involved in the development of the intestinal mucosa (Lonnerdal, 2003).
Compared to breast-fed infants, in formula-fed infants the microbiota is more diverse, containing Bacteroides, Bifidobacterium, Enterococcus, Staphylococcus, Escherichia coli, Lactobacillus and Clostridium as dominant cultivatable species. Children that are fed by breast milk but receive a supplement of formula harbour a microbiota that has overlapping microbiota characteristics of both diets (Fanaro et al., 2003). The mode of delivery is a further important factor affecting the intestinal microbiota (Fanaro et al., 2003). The microbiota of children born by caesarean section has been characterised by low numbers in Bacteroides spp. and Bifidobacterium and increased numbers of Clostridium perfringens have been found (Fanaro et al., 2003).
Oligosaccharides in human milk belong to the main components influencing the development of the intestinal microbiota (Boehm et al., 2005; Coppa et al., 2004; Newburg, 2000; Newburg et al., 2005). The complexity of human milk oligosaccharides is high and based on the variable combination of glucose, sialic acid, galactose, fucose and N-acetylglucosamine. This complex mixture is further modified by various linkages between the respective sugar residues. Since the breast milk oligosaccharides are only partially digested in the small intestine, they reach the colon and stimulate the development of a bifidogenic microbiota (Boehm et al., 2005; Coppa et al., 2004). The recognition of effects of diet constituents on the promotion of specific probiotic microorganisms has been led to the development of the concept of prebiotics by Gibson and Roberfroid (1995).
Contrary to human breast milk, standard infant formulas are virtually free of prebiotic oligosaccharides (Boehm et al., 2004, 2005). It has been proposed that standard formula-fed infants harbour a modified intestinal microbiota due to a lack of complex oligosaccharides. It was therefore aimed to supplement infant formula with prebiotic ingredients. Recent reviews summarising data of over 400 preterm and term infants, clearly demonstrate that prebiotic mixtures containing short-chain galacto-oligosaccharides and long-chain fructo-oligosaccharides stimulated the growth of Bifidobacteria and Lactobacilli, decreased faecal pH and normalised short-chain fatty acid pattern in infant stool (Boehm et al., 2004, 2005; Fanaro et al., 2005). Furthermore the presence of pathogens can be reduced to levels similar to those of breastfed infants (Boehm et al., 2004, 2005; Fanaro et al., 2005).
Preliminary data are available from clinical trials which indicate a reduced humoral allergic IgE response and reduced episodes of upper airway infection during the first year of life due to prebiotic supplementation (Boehm et al., 2005). Altogether these data show that the use of prebiotic supplements and probiotics is still in its early stages but there is promising evidence that prebiotic oligosaccharides can provide beneficial effects for formula-fed infants.
3.2.2 Prevention of necrotising enterocolitis using probiotics
Necrotising enterocolitis (NEC) is a severe inflammatory reaction of the small and large intestine with a prevalence of one to two percent of all preterm infants (Henry and Moss, 2004). Eighty percent of the affected preterm infants have a birth weight under 2000 g. Major risk factors are prematurity, early enteral feeding and unphysiological bacterial colonisation. Other predisposing factors are absence of breast milk feeding and reduced perfusion of the intestine due to artery hypotension, hypovolaemia or persistence of the ductus arteriosus Botalli.
It has been shown that intestinal microbiota of infants on intensive care units differs strongly from that of healthy neonates (Harmsen et al., 2000; Millar et al., 2003; Rubaltelli et al., 1998). The microbiota of preterm infants on intensive care units has been shown to contain potentially harmful organisms (Millar et al., 2003; Szajewska et al., 2006). Predominant bacterial species in preterm children were Enterococcus faecalis, E. coli, Enterobacter cloacae, Klebsiella pneumoniae, Staphylococcus epidermidis and Staphylococcus haemolyticus (Fanaro et al., 2003). It has been proposed that modulating the microbiota of preterm infants on intensive care units into the direction of the one found in healthy breast-fed infants may prevent necrotising enterocolitis.
Several studies attempted to colonise the gut of preterm infants by different probiotic preparations to establish a microbial environment resembling the one by breast-fed infants. Among the strains administered there were Lactobacillus rhamnosus GG (ATCC 53103), Lactobacillus acidophilus (Chris Hansen Laboratory, Milwaukee) and Bifidobacterium breve (YIT 4010). Results of the ability to colonise the intestine were variable (Kitajima et al., 1997; Reuman et al., 1986; Schultz et al., 2004).
In the recent years three high quality randomised controlled trials (RCTs) have evaluated the effect of probiotics on the prevention of necrotising enterocolitis (Bin-Nun et al., 2005; Dani et al., 2002; Lin et al., 2005). Two of them evaluated supplemented formula, one breast feeding plus dissolved probiotics administered directly by spoon. investigated probiotics strains were L. rhamnosus GG and probiotic mixtures, one containing B. infantis plus Streptococcus thermophilus plus Bifidobacterium bifidum (ABC Dophilus, Solgar, Wyeth Consumer Healthcare), the other containing L. acidophilus plus B. infantis (both ATCC 1973) (Bin-Nun et al., 2005; Dani et al., 2002; Lin et al., 2005). The study performed by Dani et al. investigating the effect of L. rhamnosus GG on the incidence of necrotising enterocolitis, bacterial sepsis and urinary tract infections included 585 preterm infants (Dani et al., 2002). Although there was a reduction in NEC and urinary tract infections the effect was not statistically significant. The two RCTs using probiotic mixtures revealed a significant decrease in incidence and severity of NEC compared to control group (1.1% versus 5.3% (Lin et al., 2005) and 4% versus 16.4% (Bin-Nun et al., 2005)).
The described RCTs did not reveal any serious side effects by colonisation with probiotic organisms (Szajewska et al., 2006). Probiotics are therefore generally considered as safe. However, it should be taken into consideration that there is a potential and real risk of administering large quantities of viable microorganisms into infants with a highly immature immune system. indigenous lactobacilli have been shown to cause rare pediatric infections in primarily immunocompromised infants (Thompson et al., 2001).
Although the results on prevention of necrotising enterocolitis are promising, no general recommendation has been given so far for the general application of probiotics in preterm children.
3.2.3 Probiotics are an effective treatment option in viral and antibiotic associated diarrhoea
Enteritis is one of the most common infectious diseases in children. Between one and four years of age most children are affected once or twice a year. Eighty to 90 % of all cases with diarrhoea in developed countries are caused by viruses. Infection with rotavirus is the most frequent, followed by adeno-and norovirus infection. Therapeutic strategies aim to avoid and treat dehydration caused by fluid loss due to watery stools and vomiting. The most effective rehydration strategy is an oral fluid administration as hypoosmolar oral rehydration solution. The World Health Organization recommends oral rehydration in children with mild and moderate dehydration. Intravenous rehydration should be restricted to children with severe dehydration. Other more recently established therapeutic approaches include modulation of intestinal fluid secretion by enkephalinase inhibitors (Salazar-Lindo et al., 2000). Successful efforts have been made to establish a safe and effective vaccination against rotavirus infection (Widdowson et al., 2005).
There is also evidence that administration of probiotic microorganisms is of significant benefit in the treatment of acute diarrhoea in infants and children. Particularly in rotaviral gastroenteritis the beneficial effect of probiotic strains on the shortening of viral diarrhoea and dehydration in children has been well documented (Allen et al., 2004; Huang et al., 2002; Szajewska et al., 2001, 2006; Van Niel et al., 2002). The following probiotic strains were among those investigated: L. rhamnosus GG, Lactobacillus reuteri (SD 2112), inactivated Lactobacillus acidophilus LB (MA 65/4 E), Bifidobacterium animalis subsp. lactis plus L. acidophilus (Infloran berna) and S. thermophilus plus Lactobacillus bulgaricus (standard starter, International Yoghurt Manufacturers Club, Paris) (Szajewska and Mrukowicz, 2001; Szajewska et al., 2006). Other nonpathogenic strains of different genera, including Escherichia, Enterococcus (LAB SF68), Streptococcus faecium (68) and Bacillus as well as the yeast Saccharomyces boulardii have also been used (Allen et al., 2004).
L. rhamnosus GG is the probiotic organism studied with the most consistent beneficial effects on acute diarrhoea in childhood when used in doses > 1010 CFUs (Szajewska and Mrukowicz, 2001). L. rhamnosus GG significantly reduced the number of watery stools in patients affected by rotavirus enteritis. Compared with placebo, the duration of diarrhoea was reduced from 13.6 to 33.6 hours with a normalised mean reduction of 20.1 hours as published in a systematic review of published randomised, double-blind, placebo-controlled trials in the field of probiotics in acute diarrhoea (Szajewska et al., 2006). There were two studies reporting decreased vomiting in the L. rhamnosus GG and L. reuteri (SD 2112) treated group (Raza et al., 1995; Shornikova et al., 1997). One study showed a reduction in diarrhoea using the non-pathogenic yeast S. boulardii (Saccharomyces cerevisiae Hansen CBS 5926) (Kurugol and Koturoglu, 2005). Preliminary data suggest that the use of E. coli Nissle 1917 may have a significant and robust effect on the duration of acute diarrhoea in children (Henker, 2007).
So far it is not clear whether bacteria used as starter cultures for milk fermentation, such as L. bulgaricus and S. thermophilus may have beneficial effects (Szajewska et al., 2006).
Although there seems to be a beneficial effect of probiotic intake on antibiotic associated diarrhoea due to C. difficile colonisation, no benefit on diarrhoea of other bacterial aetiology such as infection by Salmonella, Shigella, Campylobacter, Yersinia or Entamoeba has been proven (Guandalini et al., 2000). This lack of activity in bacterial-induced diarrhoea may be an explanation why probiotics have shown clear effects in children in developed countries (predominance of viral infections) compared to reduced efficacy in developing countries (higher incidence of bacterial infections) (Szajewska et al., 2006; Van Niel et al., 2002).
Taken together it seems to be clear that probiotics do play a beneficial therapeutic role by reducing the duration of diarrhoea and frequency of stools particularly in young children. This strain-dependent, dose-dependent effect has been shown most consistently for L. rhamnosus GG (Szajewska et al., 2006; Van Niel et al., 2002). The exact mechanism of action is not clear and involves most likely several mechanisms including anti-microbial activity, microbial competition, epithelial adherence and anti-inflammatory activities.
Another interesting and highly relevant aspect for the use of probiotics is the prevention of nosocomial diarrhoea. The incidence of nosocomial diarrhoea in children admitted to a children's hospital ranges from 4.5 to 22.6 cases per 100 admissions (Ford-Jones et al., 1990; Ponce et al., 1995; Szajewska et al., 2001). Nosocomial diarrhoea prolongs hospital stay and leads to increasing costs in medical health care. Transmission typically occurs by faecal-oral route, possibly also via droplet infection (Caul, 1994). Infants and younger children are at a higher risk to acquire nosocomial diarrhoea which is typically induced by virus infection. Bacterial infection occurs only in one percent of cases of nosocomial diarrhoea in children. Among these, C. difficile infection is the most frequent cause.
Few controlled trials in children intended to investigate the potential of probiotics for the prevention of nosocomial diarrhoea. These studies were performed using acidified formula milk supplemented with B. animalis subsp. lactis Bb12 and S. thermophilus (both Christian Hansen Laboratories Copenhagen) (Chouraqui et al., 2004; Saavedra et al., 1994) or powder containing L. rhamnosus GG (Mastretta et al., 2002; Szajewska et al., 2001). Results were partially contradictory. The initial RCT performed in 1994 evaluated infant formula supplemented with B. animalis subsp. lactis Bb12 and S. thermophilus on 55 infants in a chronic care hospital (Saavedra et al., 1994). Compared with placebo it revealed a significant reduction of the prevalence of nosocomial diarrhoea within the probiotic group (7% versus 31%) as well as shortening of rotavirus shedding. The number needed to treat was 5 (95% CI 3-20). However, if the results of the study were presented as the number of episodes per patient-month instead of numbers of episodes per patient in each group, the results between both groups would not have been significantly different (Szajewska et al., 2006). A further RCT evaluated the effect of L. rhamnosus GG on the prevention of diarrhoea (Mastretta et al., 2002). 220 children under the age of 18 months admitted to a children's hospital for other reason than infectious enteritis were included. Probiotic supplementation did not show any benefit on reduction of the incidence of diarrhoea (Chouraqui et al., 2004; Mastretta et al., 2002). In one trial L. rhamnosus GG reduced the incidence of nosocomial diarrhea and rotavirus gastroenteritis but there was no difference in the incidence of rotavirus infection (Szajewska et al., 2001).
In summary there are some promising results showing the effectiveness of L. rhamnosus GG or B. animalis subsp. lactis Bb12 plus S. thermophilus in prevention of nosocomial diarrhoea. However, lacking consistent evidence for efficacy, general advice for the use of probiotics for prevention of nosocomial diarrhoea cannot be given so far.
Antibiotic-associated diarrhoea is regarded as otherwise unexplained diarrhoea in patients treated with antibiotics (for other reasons than diarrhoea) (Bartlett, 2002). Results that show beneficial effects of probiotics in antibiotic-induced diarrhoea in adults are encouraging (Sazawal et al., 2006). There is also a moderate beneficial effect in children (Correa et al., 2005; Kotowska et al., 2005). Improving of antibiotic-associated diarrhoea has been shown for certain strains like L. rhamnosus GG, S. boulardii, B. animalis subsp. lactis Bb12 plus S. thermophilus (Szajewska et al., 2006). There is one recent RCT evaluating the effect of a commercialised formula infant formula milk containing B. animalis subsp. lactis and S. thermophilus (Nan Probiotico by Nestle Brasil) that found a significant difference in the incidence of antibiotic-associated diarrhoea in children (31% versus 16%; NNT 7; 95% CI 4-62) (Correa et al., 2005). These results suggest that probiotics could be a relevant supplement for antibiotic therapy in children but further RCTs are needed to propagate the regular use with the aim to prevent side effects of antibiotic therapy.
The potential mechanisms involved in the prevention and treatment of acute diarrhoea are discussed in more detail in Chapter 4.
3.2.4 Probiotics inhibit Helicobacter colonisation and support standard eradication therapy
Infection with the Gram-negative microaerophilic bacterium Helicobacter pylori is typically acquired in childhood. About 10% of patients develop symptoms of gastritis, peptic ulcer disease or MALT lymphoma, but anaemia and growth retardation is also associated with H. pylori colonisation in children (Czinn, 2005). The infection is a relevant health problem since approximately half of the world population is infected with H. pylori. In particular, high prevalence rates exist in developing countries and in populations with low socio-economic standard and poor hygienic conditions. H. pylori can be eradicated by anti-microbial therapy (typically amoxicillin plus clarithromycin or amoxicillin plus metronidazole) plus a proton pump inhibitor (Bourke et al., 2005; Czinn, 2005; Elitsur and Yahav, 2005). However, using the current antibiotic combinations complete eradication rate has not been achieved and current therapy is associated with gastrointestinal side effects.
There is increasing evidence that probiotics do not completely eradicate H. pylori but maintain lower levels of this pathogen in the stomach (Gotteland et al., 2006). This reduction in bacterial load seems to increase the efficacy of standard eradication therapies. Furthermore, the use of probiotics may reduce the rate of adverse effects of current standard therapies (Gotteland et al., 2006).
School children from a low socioeconomic area of Santiago de Chile that received one month live Lactobacillus johnsonii La1 (but not L. paracasei ST11 or heat-killed L. johnsonii La1) showed moderately reduced DOB values in the (13)C-urea breath test, indicating a direct modulation in H. pylori colonisation by living L. johnsonii La1 (Cruchet et al., 2003).
H. pylori infection and probiotic interventions in adults are discussed in Chapter 4.
3.2.5 Probiotics and the prevention of allergy in children
According to the hygiene hypothesis the increasing number of allergic diseases in western countries is caused by reduced exposure to pathogens in early infancy, improved hygiene standards, changes in diet, different delivery mode and smaller family sizes (Schaub et al., 2006; Strachan, 1989).
The analysis of the composition of the faecal microbiota from healthy and allergic children found in affected children higher amounts of the adult type B. adolescentis compared with healthy infants harbouring more B. bifidum. It was suggested that B. bifidum has greater adhesive qualities which may help to stabilise the mucosal barrier and prevent adsorption of antigenic proteins (He et al., 2001).
Similarly, the intestinal microbiota of atopic children has been found to be different from non-atopic individuals. The bacterial cellular fatty acid profile in stools from atopic children was significantly different from the one in healthy children. These findings correlated with more clostridia and a lower content of bifidobacteria in the faeces of allergic individuals (Bjorksten et al., 2001; Kalliomaki et al., 2001a). This might be a strong indication of an interaction of the gut microbiota with the immune system influencing the onset of allergy and atopy.
Kalliomaki et al. investigated the long term effect of early colonisation of infants with probiotic bacteria on allergy prevention. one hundred and fifty-nine pregnant women who had a positive family history for atopy were supplemented during their last month of pregnancy with L. rhamnosus GG. Probiotic administration was continued in mothers and children for six months after delivery. The primary end point was chronic atopic eczema. Probiotic treatment led to a significant reduction in the prevalence in at-risk infants at the age of two years (46 % versus 23 %) (Kalliomaki et al., 2001b). However, there was no decrease in antigen-specific IgE by L. rhamnosus GG administration. The authors performed a four-year follow up of the study group. Sixty-seven percent of the initially randomised children were re-examined. In the probiotic supplemented population there was a significantly decreased prevalence of atopic eczema compared with the non-treated group (14 of 53 versus 25 of 54 children) (Kalliomaki et al., 2003). No difference was found upon skin prick test reactivity. Another long-term follow up study was able to shown that intentional colonisation of the intestine with E. coli 083 (K24LH31) after birth decreased the incidence of allergies after 10 and 20 years after colonisation (Lodinova-Zadnikova et al., 2003).
Results for the use of probiotic organisms in the field of allergy prevention are encouraging. Effects on early intervention seem to lead to longer lasting effects.
3.2.6 Developing probiotic and prebiotic products for infants -potential and open questions
The development of probiotics for use in children is a challenging task. Due to the flexibility of the developing microbiota in early childhood there is a clear potential to influence and modulate the intestinal microbiota to achieve health benefits. Furthermore there are several defined applications where probiotics have been shown therapeutic benefit. The therapeutic applications include relevant disorders with high incidence like infectious enteritis, H. pylori infection or allergy. Already these few applications mean that probiotics could be relevant for potentially every child.
However, there are several risks and questions that have not been answered yet: there is still some uncertainty about the safety of probiotics in very young children with an immature immune system as addressed earlier. In fact, there are few reports of sepsis in patients treated with L. rhamnosus GG. A 6-week-old full-term baby with double outlet right ventricle and a 6-year-old child with jejunostomy and several infections are among those reported (Land et al., 2005). There are also reports of fungaemia by S. boulardii (Saccharomyces cerevisiae Hansen CBS 5926) after therapeutic use as probiotic (Hennequin et al., 2000; Herbrecht and Nivoix, 2005; Rijnders et al., 2000). This was probably due to contamination of central venous catheters while opening capsules for enteral feeding. Very restricted information is available about the effects and side effects of probiotic use in the long-term administration over several years. To our knowledge there is only one published report spanning a two-decade post-treatment follow up of therapeutic modulation of the intestinal microbiota by postnatal administration of E. coli 083 (K24LH31) (Lodinova-Zadnikova et al., 2003).
Most studies performed on children used a specific therapeutic formulation of the probiotic microorganisms. Only in rare cases have dairy products been used to investigate health effects in RCTs. However, the route of delivery is crucial for probiotic microorganisms. It is therefore not directly possible to extrapolate from effects seen in a specific pharmaceutical formulation with a defined dosage, to judge effects of probiotic bacterial strains used as dietary dairy product supplements. The viability of probiotics varies in different commercial products ready for consumption. There is a large variability of therapeutic effects between different strains and even quite related strains do not necessarily exhibit similar effects. Therapeutic effects cannot be generalised and are therefore restricted on a certain probiotic strain tested for a defined indication. However, data that enable the comparison of different probiotics side by side are limited. Since study design and populations vary between different RCTs, comparison between studies is error prone and further studies are needed to find the most efficient strains.
Prebiotics are easier to manage in industrialised processes compared to living probiotics. However, effects in terms of modulation of the microbiota are difficult to study and no relevant surrogate markers are useful for judging health effects. More randomised controlled clinical trials are needed to study the effects of prebiotics on disease prevention and treatment.
The mass market for supplementation of dietary products with probiotics carries some risk of transmission of pathogenicity or antibiotic resistance between different bacteria and the probiotics used. This would be in particularly relevant when genetically modified bacteria enter the use in humans.
Although there is an increasing number of well designed RCTs indicating the benefits of probiotics, we still do not understand the mechanisms of their biological activity in detail. Consistent with this, the influence of long-term administration of probiotics on allergy, autoimmune diseases, and infection control has not been understood.
Non-pathogenic probiotic bacteria may become a relevant vehicle for the delivery of drugs as has been shown for IL-10 producing Lactococcus lactis. IL-10 secreting genetically modified lactococci were able reduce colitis in a mouse model (Steidler et al., 2000). Strategies have been applied to use genetically modified probiotics for vaccination.
Probiotic organisms start to enter the food mass market as well as the paediatric medical routine. To address the open questions seems to be a precondition for the use of probiotics as therapeutics, to meet safety requirements of probiotics and diary product in children and adolescents as well as to maintain patients and consumer confidence.
Was this article helpful?