Developing infant gut microflora and complementary nutrition

ABSTRACT

Human infants are born with sparse microflora in their gastrointestinal tracts. Acquisition of pioneer bacteria and their initial colonisation are affected by a number of factors, including mode of birth, feeding practices, exposure to antibiotics, and environmental conditions. Subsequent diversification and development of the microbiota continues dynamically until it reaches maturity when the infant is around 3 years old. An important window of opportunity to affect the structure of developing microflora is likely to coincide with the weaning period when a gradual transition in diet from milk-based infant feeding to first family-based solid foods occurs. Although the ability of many dietary factors to modulate the composition of gastrointestinal microflora is well established, it is not so well understood in children. Moreover, there are cultural, economic, varying feeding practices, and geographical factors that play roles in the choice of complementary foods in different regions of the world. We provide a review of recent literature on development of infant microflora and how complementary feeding practices from different cultures may affect the infant gut microbiota during early childhood.

KEYWORDS:

• Infant microbiome
• complementary foods
• breast feeding
• human milk oligosaccharides
• plant based foods and prebiotics

Introduction

Often regarded as a ‘living organ’ on its own right, the human gut microbiome consists of microbial entities connected by a highly complex network of microbial competition and cooperation. The intimate relationship between the structure and function of the human gut microbiome and health is now well recognised. Impaired acquisition of pioneer bacteria, followed by a reduction in microbial diversity and the ensuing gut dysbiosis, have been shown to be associated with many childhood and later life disorders (Hooper et al. 2012; Tremaroli and Backhed 2012; Jain and Walker 2015; Tang and Hazen 2017; Yoshida et al. 2018; Pulikkan et al. 2019). Our understanding of the role of human gut microbiota in human physiology and health has made rapid advances in the recent years as a result of the emergence of molecular ‘tool box’ and multiple integrative ‘omics’ technologies, including metagenomics, metabolomics and transcriptomics. The crucial role that microbiota play in shaping the immune system and the digestive functions of developing infants is now well recognised (Macpherson and Harris 2004; Purchiaroni et al. 2013; Gomez de Aguero et al. 2016; Macpherson et al. 2017; Robertson et al. 2019). In addition to other factors, including mode of birth (Dominguez-Bello et al. 2010; Perez-Munoz et al. 2017) and host genetics (Spor et al. 2011; Bonder et al. 2016), diet is one of the strongest environmental factors that determines the diversity of infant gut microbiota. Differences in early feeding practices such as breastfeeding versus formula-feeding have an impact on determining the initial bacterial composition and the alpha diversity of the infant gut microbiome (Penders et al. 2006; Chong et al. 2018,). Profound and sustained changes occur to the community structure of the infant gut microbiome with the introduction of solid foods (Fallani et al. 2011; Koenig et al. 2011; Stewart et al. 2018). During this period of dietary transition, the microbial composition of the infant gut continuously and flexibly adapts to the introduced dietary components, acquiring new bacterial strains and establishing dominant species that have the ability to utilise these new substrates for growth (Backhed et al. 2015). This period represents a critical window of opportunity to influence the community structure of a developing infant’s microbiome via dietary interventions. Parental choices for first foods thus may have a significant impact on shaping the structure and diversity of infants’ developing gut microbiota, leading to better health outcomes later in life (Movassagh et al. 2017). In this review we summarise current knowledge on the developmental trajectory of the infant microbiome and the role that early nutrition plays in its healthy development and maturation.

Initial acquisition of gut bacteria

There has been a long-standing belief that the anatomical, physiological, and immunological features of placenta ensure that prenatal developments of a foetus occur in a sterile environment, and the first microbes are acquired from the mother during the birthing process. However, this view may not be accurate, because the presence of microorganisms in meconium, placenta, and amniotic fluid has been reported in several studies (Jimenez et al. 2008; Aagaard et al. 2014; Blaser and Dominguez-Bello 2016; Collado et al. 2016). While these observations suggest that microbial colonisation may begin in utero preceding birth, such claims still lack direct and unequivocal evidence (Perez-Munoz et al. 2017; Robertson et al. 2019). A vertical transfer of microbiota from mother to the new-born is a much more likely way of initial acquisition of microbes during birth. The establishment of dominant bacterial types often depends on the mode of birth and other environmental factors. During a vaginal birth, the baby acquires its bacteria from its mother, and the mother’s gut microbiome appears to be the biggest influencer of the infant’s gut microbiome (Ferretti et al. 2018). In case of a caesarean delivery, however, the mother-to-baby transmission pattern is disrupted. The more common bacteria that are found in babies born by caesarean section are reported to be bacteria associated with a hospital setting (Shao et al. 2019). The long-term differences in the composition of gut microbiota of babies born via caesarean section in comparison to vaginal birth have been documented in many studies, where the microbiome of the former has been generally found to be less diverse than that of the latter (Jakobsson et al. 2014; Backhed et al. 2015). Interestingly in another recent study, Chu et al. (2017) assessed the impact of mode of delivery on the maturation of infant microbiome community structure and function across multiple body sites. Their study concluded that within the first 6 weeks of life, the infant microbiota undergoes significant reorganisation that is primarily driven by body site and not by mode of delivery.

Although there is little doubt about the occurrence of vertical transmission of bacterial strains from mother to infant (Backhed et al. 2015; Asnicar et al. 2017; Sakwinska et al. 2017), there is still uncertainty about which maternal body site serves as the primary source of the initial bacterial inoculum that colonises and subsequently dominates the early infant gut. Some studies found little similarity between the bacterial taxa present in infant faecal samples and those present in maternal vaginal samples (Sakwinska et al. 2017). This suggests that vaginal inoculum may not be the source of bacteria that succeed in colonising and becoming the initial dominant species in the infant gut. A study by Nayfach et al. (2016) used molecular techniques and sophisticated bioinformatics to track the transmission of specific bacterial taxa from mother to infant in a precise manner. They analysed the faecal metagenomes of 96 Swedish mothers and their infants collected over more than a year, and showed that the early colonising bacterial species transmitted to the infant originated from the maternal gastrointestinal tract, whereas late colonisers were often transmitted from other sources in the external environmental.

Typically, faecal samples of a new-born are enriched with facultative anaerobic bacteria belonging to genera such as Enterococcus, Escherichia/Shigella, Streptococcus and Staphylococcus (Iozzo and Sanguinetti 2018). These facultative anaerobes consume oxygen to slowly create habitable anaerobic environments (Arrieta et al. 2014; Backhed et al. 2015), priming the gut for the subsequent colonisation by more strict anaerobes including those of the genera Bifidobacterium and Clostridium.

The composition of breast milk also plays an important role in facilitating the selection of early colonisers, particularly the human milk oligosaccharides (HMOs). More than 200 structurally diverse unconjugated 3–15 carbon unit long oligosaccharides consisting of the monosaccharides glucose, galactose, N-acetylglucosamine (GlcNAc), fucose, and N-acetylneuramic acid (sialic acid) are present in human breast milk (Ayechu-Muruzabal et al. 2018). Bifidobacterial genomes, especially Bifidobacterium longum subsp. infantis, show a high degree of specialisation for utilising HMOs (Underwood et al. 2015), often encoding gene clusters consisting of numerous Carbohydrate-Active enZymes (CAZymes), ABC transporters, and solute-binding proteins that specifically target oligosaccharides derived from HMOs (Thomson et al. 2018), thus enabling B. longum subsp. infantis to competitively colonise the gut of breast-fed infants. One of the reasons that many commercial infant formulae are supplemented with non-digestible oligosaccharides such as galacto-oligosaccharides and fructo-oligosaccharides is to partially mimic the prebiotic properties of human breast milk. Although the dominance of bifidobacteria in the infant gut is often considered an indication of normal gut maturation, not all infants seem to reach this developmental milestone. In some studies, high abundances of Proteobacteria and Firmicutes, and very low numbers of Actinobacteria – regardless of the birth modes – have been reported in healthy neonates and infants from China and Brazil (Kuang et al. 2016). Similarly in other studies, Bacteroides rather than bifidobacteria were reported to be the dominant group of bacteria in breast-fed babies (Tannock et al. 2016; de Muinck and Trosvik 2018). It is known that certain Bacteroides can degrade HMOs via mucin utilisation pathways (Marcobal et al. 2011), which may explain this observation.

In a recent review Korpela and de Vos (2018) provided a global view of the temporal progression of microbiota development in the infant gut during early life. They looked at cumulative data from more than 34 studies conducted in different geographic regions of the world and analysed age-associated changes in the relative abundance of five of the most abundant classes of bacteria in the infant gut. Figure 1 illustrates an age-adjusted evaluation of the average relative abundances in the infant gut from birth to 24 months of age using a subset of the microbiome metadata compiled by Korpela and de Vos. Metadata from 29 studies were used in Figure 1 (data points from infants who had been given antibiotics or who were over the age of 24 months were excluded). Temporal changes in the infant gut microbial populations of key bacterial groups can be clearly observed from these data. Initial colonising bacteria are bacilli (aerobes and facultative aerobes), which give way to colonisation by strict anaerobes such as bifidobacteria and Bacteroides. Bacterial taxa such as Proteobacteria that initially colonise the new-born gut gradually make way for the colonisation of bifidobacteria and Bacteroides and as the complexity of diet increases, there is an increase in relative abundance of Clostridia. While variables such as sample origin (country), birth mode and feeding practices are acknowledged as influencers of the microbial landscape, the age and maturation of the infant gut also influence the microbial colonisation patterns.

Figure 1. Average relative abundance of the five most abundant bacterial classes from the early life microbiota studies (N = 29), as a subset from the data compiled by Korpela and de Vos (2018). Studies with infants who had received antibiotics or those who were over 24 months of age were excluded. The trend lines were calculated using a loess smoother on the square-root of age, with each study weighted by its sample size. (Adapted from the original figure published in Current Opinion in Microbiology 44: 70–78, under Elsevier Copyright license# 4647300186567, obtained on 13 August 2019.)

Given that the postnatal development and assembly of the gut microbial community is associated with long-term health benefits, it would be helpful if some microbial markers of normal and healthy progression of gut microbiota could be identified. One approach that has been suggested and used in the literature is description of age-discriminatory bacterial taxa. The changes in relative abundance of these taxa provide a microbial signature that describes a programme of normal gut microbiota development shared by biologically unrelated healthy infants/children from birth to 3 years of age (Subramanian et al. 2014).

Current infant complementary feeding practices

The World Health Organization (WHO) guideline recommends ‘exclusive breastfeeding for the first 6-months of age and continuation for up to 2-years of age and beyond’ (https://www.who.int/nutrition/topics/exclusive_breastfeeding/en/). Breastmilk provides sufficient energy and nutrients up until 6 months of age, beyond which breastfeeding alone becomes increasingly insufficient to provide appropriate energy density and essential nutrients for a growing child. Complementary foods therefore must accompany and gradually replace breast milk not only to provide essential micro- and macronutrients for growth but also to promote long-term health outcomes. There is no consensus on the timing of introduction of complementary foods; timing and nature of foods to be given vary between countries depending on cultural influences and food habits. Practical challenges of having accessibility to fresh foods and affordability of fortified commercial food products can also affect food choices. In developing African and South Asian countries, fewer than 15% of children have access to the supply of energy-dense meats (Ayana et al. 2017) to replace breast milk as the source of protein and fat after breast feeding ceases. Animal-sourced foods are high in many macronutrients (protein) and micronutrients (vitamin A, vitamin B12, riboflavin, calcium, iron, and zinc) that have high bioavailability (Murphy and Allen 2003). Many children (especially in lower socio-economic groups) in developing countries where meat consumption is low are often deficient in vitamin B12 (Murphy and Allen 2003). Monotonous cereal- and starch-based diets are often low in essential micronutrients such as calcium, iron and zinc and fall short of complementing breast milk (Krebs et al. 2013; Osendarp et al. 2016). Although consumption of animal-sourced foods is often associated with a high quality diet and lean tissue mass gain (Robinson et al. 2009), many children growing up in developing countries lack regular intakes of animal-sourced foods, and hence must rely on plant-based diets, especially fruits and vegetables rich in vitamin A and vitamin C, for essential nutrition. However, statistics reported that children growing up both in developing and developed countries lack regular intakes of fruits and vegetables. Globally, only 29.4% of children meet the minimum dietary diversity (White et al. 2017). In Ethiopia, only 23.7% of children have acceptable dietary diversity, with the major dietary sources being cereals and legumes, while intakes of vegetable-, fruit-, and animal-sourced foods are very low (Baye et al. 2012). In India, minimum dietary diversity among children aged 6–23 months was reported as 15% (Senarath et al. 2012), with only 31.5% of 6- to 23-month-old children consuming vitamin A-rich fruits and vegetables (White et al. 2017). In China, cereal rice, porridge, dumplings, whole eggs, meat, vegetables, and fruits were dominant choices of complementary foods (Li et al. 2003). Porridge, which was reported to be the major complementary food for infants in China and South Asia, is low in nutrient density and needs to be consumed in a large volume to fulfil the high energy requirements of growing infants (Dewey 2016). In socio-economically challenged settings, commercially available fortified foods are often unaffordable, and fortified foods may have limited benefits since infants and children often do not consume enough food for fortification to have an impact (Dutta et al. 2006). In the USA, up to 27% of infants aged 6 to18 months consume no vegetables, and up to 23% consume no fruits on any given day (Siega-Riz et al. 2010). Reduced vegetable/fruit intake is partly associated with consumption of commercially prepared infant foods that are often high in salt and sugar (Campoy et al. 2018). Specific individual complimentary foods used in most European countries generally meet or exceed dietary requirements for energy, macronutrients and micronutrients, although in some subgroups, the intake of specific nutrients such as ω-3 polyunsaturated fatty acids (PUFAs), vitamin D, or iodine is borderline (EFSA NDA Panel 2013). There are several opinions regarding the right age for introduction of complementary foods and if attention should also be paid to taste and texture. It has been suggested that a prolonged use of pureed foods should be avoided and by 8 months of age finger foods such as soft cooked vegetable slices should be introduced (Marduel and Vernet 2018). The European Society for Paediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) Committee on Nutrition (CoN) suggests that 9–10 months of age is the critical window for introducing lumpy solid foods, and its introduction in later age is related to a high risk of feeding difficulties and unhealthy eating habits (Agostoni et al. 2008; Mennella and Trabulsi 2012; Fewtrell et al. 2017).

Complementary foods and their interaction with microbiome

With the introduction of complementary foods, the primary source of dietary carbohydrates changes from lactose and oligosaccharides present in breast milk or infant formulae to plant-derived polysaccharides. As discussed earlier, the microbiome of breast-fed babies is often dominated by HMO-utilising Bifidobacterium spp. (e.g. B. longum subsp. infantis, Bifidobacterium bifidum, and Bifidobacterium breve), (Tannock et al. 2013; Bergstrom et al. 2014). Compared with HMOs, the composition and chemical structures of plant-derived carbohydrates are highly diverse, resulting in selection of microbes that can utilise these complex substrates for growth. Among the members of the genus Bifidobacterium, B. breve, B. longum subsp. longum, B. adolescentis, and B. catenulatum are known to ferment mucin as well as a wide range of plant-derived oligosaccharides and short-chain sugars including malto-, fructo-, xylo, and galacto-oligosaccharides (Pokusaeva et al. 2011), and therefore they can readily adapt to the changing diet, and persist. Some members of Lachnospiraceae, Ruminococcaceae, Bacteroidaceae, and Prevotellaceae families are particularly effective in fermenting both insoluble and soluble dietary fibres and crystalline starch granules (Flint and Bayer 2008). These ‘primary degrading’ taxa of infant gut bacteria essentially initiate the process of carbohydrate utilisation and orchestrate the expansion of the microbial diversity by releasing a variety of oligomeric degradation products for neighbouring gut bacteria that are less well adapted for degradation of complex carbohydrates (Flint et al. 2012 ). Among dietary fibres, pectin, hemicellulose, and resistant starch are particularly relevant to infant nutrition, as these are typically abundant in first foods based on cereals, fruits and vegetables, including baby cereals produced with oatmeal, rice and barley, mashed or pureed fruits and vegetables, cooked and mashed starchy vegetables, beans, and legumes. In a recent study Gopalsamy et al. (2019) examined whether high amylose maize starch (a resistant starch and a known prebiotic for adults) could be fermented by the developing microbiota of infants. They collected faecal samples from 17 infants at two time points, pre-weaning and within 10 weeks of the introduction of first solids. Fermentation of resistant starch and the associated changes in the composition of microflora were assessed using in vitro batch fermentation at 24 h. This study showed that independently of weaning stage, infant faecal inocula were capable of fermenting resistant starch. Furthermore, the introduction of solids not only resulted in increasing the fermentation capacity of inocula but also increased the Shannon diversity index with increased abundance of Bifidobacterium and Bacteroides. Many other known bacteria including Faecalibacterium prausnitzii, Eubacterium rectale, and Ruminococcus bromii, and many unassigned members of Clostridiales are known to degrade complex carbohydrates and produce short chain fatty acids (SCFAs) such as butyrate and propionate (Maier et al. 2017). In an infant gut, F. prausnitzii, Roseburia spp., and Eubacterium hallii (Bergstrom et al. 2014) start appearing at around 12 months of age, which often coincides with the introduction of more diverse dietary carbohydrates. Members of the families Ruminococcaceae and Lachnospiraceae, and Bacteroides spp. may (Yassour et al. 2018) or may not be present before introduction of solid foods (Koenig et al. 2011), but they eventually increase in abundance with the addition of solid foods into the infant diet (Laursen et al. 2017). The degradation by-products of complex carbohydrates allow a rapid expansion of secondary bacterial populations that perform essential functions such as vitamin synthesis (Yatsunenko et al. 2012; Backhed et al. 2015) and SCFA synthesis (Roy et al. 2006). It is important to consider that pectic polysaccharides constitute one of the major complex carbohydrate components of plant cell walls, and have been shown to stimulate the colonic production of butyrate, possibly through utilisation of pectic oligosaccharides by butyrate-producing bacteria such as F. prausnitizii (Lopez-Siles et al. 2012; Chung et al. 2016). F. prausnitizii is one of the most abundant gut bacteria and plays an important role in providing butyric energy to the colonocytes. Together with hemicellulose-rich whole grains and bran xyloglucan and arabinoxylan, the most abundantly present forms of hemicellulosic carbohydrates have been shown to promote acetate, propionate, and butyrate production (Rumpagaporn et al. 2015; Brahma et al. 2017; Huang et al. 2019). In addition, butyrogenic species that belong to the family Lachnospiraceae, whose members include Roseburia intestinalis and E. rectale, also use arabinoxylan for growth (Sheridan et al. 2016). Thus early complementary food choices have significant potential to affect the community structure of the developing infant microbiome.

The influence of dietary carbohydrates on the composition of the developing microbiome of children has been demonstrated in several studies. By comparing cohorts of children, de Filippo et al. (2010) showed that the microbiome of children living in Burkina Faso who consumed diets rich in complex carbohydrates from cereal and legume sources was more enriched in Bacteroidetes than that of children living in Italy (57% vs 22%). In addition, diets enriched with complex carbohydrates also resulted in increased abundance of several specialised starch-degrading bacterial taxa such as Prevotella and Xylanibacter in Burkina Faso children, which were not found in the Italian cohort. Instead, Italian children had microbiomes enriched in Firmicutes (63% vs 27%), which was probably driven by their fibre-poor, but calorie-rich diet. These differences in the glycan-degrading capacities of Burkina Faso and Italian cohorts’ microbiota showed a strong correlation to high dietary glycan consumption. It is worth pointing here that the two childhood cohorts used in this study also differed in their environment therefore environmental factors also played a role in the observed differences in the gut microbiota.

In another study, Malawian, Amerindian, and rural African children whose main diets consisted of maize, corn, cassava, and locally produced cereals, grains, and vegetables showed higher microbial diversity and similarity than their US and European counterparts who consumed a modern Western diet (Yatsunenko et al. 2012). Despite the large inter-individual and chronological fluctuations observed in the human gut microbial composition, studies showed that the diet-driven functional microbial core probably remains resilient to dietary alterations for an extended period of time, possibly as an adaptation mechanism to provide a safety buffer against unexpected dietary events such as starvation or overeating (Thaiss et al. 2016). An example of microbiome stabilisation was observed among samples obtained from acutely malnourished children living in sub-Saharan Africa and South Asia, who suffered from micronutrient deficiencies, compromised immune systems, increased susceptibility to infection and diarrhoea, were underweight, stunted, and wasted, which together heavily contributed to mortality and long-lasting disabilities observed in these children (Black et al. 2008). Studies showed that the faecal microbiome of Bangladeshi and Malawian children (<3 years of age) diagnosed with severe acute malnutrition were immature, and deviated from the age-discriminatory regression models derived from the microbiome of healthy children (Smith et al. 2013; Subramanian et al. 2014). Short-term (1 month) administration of commercial therapeutic foods supplemented with micronutrients did not promote sustainable changes in the microbiome configuration before, during and after treatment, with the microbiota composition remaining persistently stagnated compared with that of healthy groups, and the children mostly remained chronically underweight >4 months after treatment (Smith et al. 2013; Subramanian et al. 2014). There is growing evidence that the gut microbiome and host genetics are largely independent, and hereditary transmission of microbiota is minimal except during the neonatal period (Yatsunenko et al. 2012; Rothschild et al. 2018). Lifestyle-changing long-term intervention measures encompassing diet, geography (He et al. 2018), and sanitation have emerged as important adjustable parameters that may determine microbiome stability and dysbiosis leading to disease susceptibility, and young children are not exceptions to these rules.

Concluding remarks

There are many factors that influence this gradual temporal maturation of the infant gut microbiome. The major factors include age, birth mode, early feeding practices, geography, immediate microbial environment, exposure to antibiotics, and complementary foods at the time of weaning. While many of these factors are immutable, the choice of complementary foods offers an attractive opportunity to modulate and influence the developing infant gut microbiome, although the use of these foods depends on cultural and geographical influences, palatability, consumer acceptance and availability.

Acknowledgements

We acknowledge Dr Duncan Hedderley for his expert advice with Figure 1.

Disclosure statement

No potential conflict of interest was reported by the author(s).

ORCID

Caroline C. Kim http://orcid.org/0000-0003-3393-8364

Shanthi G. Parkar http://orcid.org/0000-0002-5737-7867

Pramod K. Gopal http://orcid.org/0000-0002-3434-5555