The first 1,000 days: how a baby’s microbiome takes shape

A child’s gut microbes arrive mainly at birth, and the microbiome takes shape over the first few years. The biggest influences are breastfeeding, the way the baby is born, antibiotics and the environment the child grows up in. In a large study of 903 children, gut microbes went through three phases and settled down at around two and a half years of age.

Does a baby have microbes in the womb?

Most likely not. Over the past decade some studies found traces of microbes in the placenta and the amniotic fluid, but a major 2023 analysis concluded that these were most likely contamination during sampling or during DNA testing.

The authors have one more argument. If microbes lived in a healthy foetus, scientists could not raise animals that are completely free of microbes, and they have been doing so for decades. So colonisation by microbes starts at birth.

How does the way a baby is born affect its microbes?

Strongly in the first weeks, and much less after the first year. In a vaginal birth the baby receives bacteria from its mother, for example from the genus Bacteroides. In a large British study of 596 babies, this transfer was disrupted after a caesarean section, and bacteria from the hospital environment, such as Enterococcus, Enterobacter and Klebsiella, settled in the gut more often.

The same was seen, to a lesser extent, in babies whose mothers received antibiotics during labour and in babies who were not breastfed in their first month. And in the study of 903 children, Bacteroides went with greater diversity and faster maturation of the microbiome.

By the first year the differences fade in most cases. In a Danish study of 700 children, asthma by the age of 6 was more common only in those born by caesarean whose microbes at 1 year still carried the “imprint” of the caesarean. A caesarean is often necessary and is not “bad” in itself.

Why does breast milk feed the bacteria, not just the baby?

Because it contains sugars the baby cannot digest. The oligosaccharides in breast milk are its third largest solid component after lactose and fat, usually 9 to 24 grams per litre, more than all of its proteins. More than 200 kinds are known, but 20–25 of them make up over 95% of the total.

These sugars reach the gut intact, where they are broken down mainly by bifidobacteria. In the study of 903 children, breastfeeding was the strongest factor in the make-up of the microbes: breastfed babies had more Bifidobacterium breve and B. bifidum, and once breast milk stopped, the microbiome “matured” faster.

Bifidobacteria probably also train the immune system. A 2021 study links their absence, and especially the absence of their genes for breaking down the sugars in breast milk, with more inflammation in babies’ bodies in the first months. The authors note that these bacteria are becoming increasingly rare in the modern world. How the gut and the immune system also affect the skin is covered in our article on the gut–skin axis.

Formula reproduces the nutritional composition of breast milk, but not the microbes in it. That does not mean formula-fed children are “problematic”. Whether a baby needs any supplement is for the paediatrician to decide.

When does a child’s microbiome settle down?

At around two and a half years. In the American and European TEDDY study, scientists analysed more than 12,000 stool samples from 903 children aged 3 to 46 months. Up to month 14 the microbiome develops fastest, between months 15 and 30 it is in transition, and from month 31 onwards it is stable.

Stable does not mean finished, though. A Swedish study that followed 471 children to the age of 5 shows that some microbes arrive late and at 5 years have still not reached adult levels. These include the archaea Methanobrevibacter, microbes from a separate kingdom, and bacteria of the family Christensenellaceae. It is exactly these “late arrivals” that go with greater diversity, in both children and adults.

How do antibiotics affect a child’s microbes?

They reduce their diversity, sometimes for a long time. A 2024 systematic review of 89 studies with 9,712 children shows that all the antibiotics studied reduced the diversity of gut microbes. After penicillins the drop can last 6–12 months, after macrolides it is still visible at the last check-up, 12–24 months later, and after an antibiotic in the first month of life diversity is lower even at 3 years.

Antibiotics also reduce beneficial bifidobacteria and lactobacilli, and often increase bacteria of the family Enterobacteriaceae other than E. coli. Macrolides have the strongest effect and the combination trimethoprim/sulfamethoxazole the weakest.

This does not mean a child should not take an antibiotic when one is needed. It means it is best taken only when the doctor decides it is necessary. What helps the gut after a course is covered in our article on how the microbiome recovers after antibiotics.

Do farms and pets help?

It seems so. The Amish and the Hutterites in the US are two communities with similar genetic ancestry and ways of life. The Amish, however, live on small family farms with cows and horses, while the Hutterites work on large industrialised farms. Asthma affects 5.2% of Amish children and 21.3% of Hutterite children.

The dust in Amish homes contains 6.8 times more endotoxins, substances from the outer layer of bacteria. When scientists gave mice an extract of this dust, it suppressed the signs of allergic asthma, while dust from Hutterite homes did not. The Amish environment seems to “train” the children’s innate immunity.

Pets matter too. In the TEDDY study, the microbiome of children living with brothers and sisters or with a dog or cat matured faster.

Does a baby need a probiotic?

Only if there is a specific reason. The European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) recommends particular strains for a few conditions: acute diarrhoea, preventing diarrhoea during antibiotics, hospital-acquired diarrhoea, necrotising enterocolitis in premature babies, treating H. pylori, functional abdominal pain and infant colic. The society advises against giving probiotics with no documented health benefit.

Probiotics are not interchangeable, because the effect depends on the specific strain. That is why the paediatrician decides whether a baby needs a probiotic and which one.

What helps a child’s microbiome?

1. 1Breast milk, if possible. It feeds the bifidobacteria and is the strongest factor for the microbes in the first year.
2. 2Antibiotics only when the doctor decides they are needed. In babies their effect on the microbes lasts for months.
3. 3Varied food from six months on, with fruit, vegetables, legumes and wholegrains.
4. 4Contact with animals and nature. Children who grow up with a dog, a cat, brothers and sisters or on a farm meet more microbes.
5. 5Patience. The microbiome keeps taking shape for years, so diet and environment still matter after the age of two.

What does the BugSpeaks gut test show in older children?

In children aged 6 and over, the gut test shows 10 things related to the topics of this article, from bifidobacteria and diversity to the traces antibiotics leave in the gut. Here they are, as they appear in the report:

• ·“Probiotic Characterization”: which of 42 probiotic species are in the gut and how much, including 7 species of bifidobacteria, such as Bifidobacterium longum, B. breve, B. bifidum and B. infantis, and 21 lactobacilli, on a 7-zone scale. Breast milk feeds exactly the bifidobacteria, and antibiotics reduce both them and the lactobacilli.
• ·“Probiotic Supplementation Summary”: for each of the 42 species, the report notes whether to consider a supplement, carry on as before or avoid one. European paediatric gastroenterologists recommend specific strains for specific conditions, so it helps the paediatrician to know what the child already has.
• ·“Shannon Diversity”: how diverse the microbes are compared with reference groups of healthy people from different parts of the world, on a 3-zone scale. Diversity keeps changing until about two and a half years, and after an antibiotic in the first month of life it is lower even at 3 years.
• ·“Foundation Microbiota”: the share of 36 species, including 7 species of Bacteroides, Faecalibacterium prausnitzii and Bifidobacterium longum, on a 7-zone scale. After a caesarean the transfer of Bacteroides from the mother is disrupted, and these bacteria go with greater diversity and faster maturation of the microbiome.
• ·“Kingdom Distribution” and “Top 5 Abundant Species”: what share of the microbes are bacteria, archaea, fungi, eukaryotes and viruses, and which species dominate each kingdom. The archaea Methanobrevibacter arrive late and at 5 years have still not reached adult levels.
• ·“Antibiotic Recovery Potential”: an estimate of how well the microbiome would recover after a course of antibiotics. In children the drop in diversity after penicillins can last 6–12 months, and after macrolides even longer.
• ·“Antibiotic Resistance”: which known resistance genes against 90 antibiotics and antimicrobial substances the gut bacteria carry. Hospital bacteria, which settle more often in the guts of babies born by caesarean, also carry resistance genes. This is general information about the microbiome, not an antibiogram, and it is not a guide for choosing an antibiotic.
• ·“Opportunistic microbes”: the share of 52 species compared with healthy people, including Enterococcus faecalis and Klebsiella pneumoniae. Bacteria from these genera settle more often after a caesarean, and after antibiotics bacteria of the family Enterobacteriaceae often increase in children. This is information about the balance of the microbes, not a test for infection.
• ·“Potential Gut Inflammation” in the “Health Indicators” section: whether microbes linked to inflammation dominate the gut and whether there are enough of the helpful ones. In babies, a lack of bifidobacteria goes with more inflammation in the body.
• ·“Nutrition Recommendations”: a 12-week plan in 3 phases with about 250 foods in 11 groups, each with a label showing how often to eat it. Before using it with a child, discuss it with the paediatrician.

We do not recommend the test for children under 6 unless the child’s doctor recommends or approves it, because until then the microbiome is still forming and changes quickly. A test for a child is ordered by a parent or guardian. This is information about the microbes, not an assessment of the child’s development or health, and not a diagnosis, so show the report to the paediatrician.

In short:

• ·Colonisation by microbes starts mainly at birth. Traces of microbes in the womb are most likely contamination of the samples.
• ·Breast milk contains 9 to 24 grams per litre of oligosaccharides that the baby does not digest, and they feed the bifidobacteria.
• ·A child’s microbiome goes through three phases and settles at about two and a half years, but some microbes have still not reached adult levels at 5.
• ·Antibiotics reduce the diversity of children’s microbes for months, and after an antibiotic in the first month of life it is lower even at 3 years.
• ·In children aged 6 and over, the BugSpeaks gut test shows 10 things related to this article, from bifidobacteria to resistance genes.

This article is for educational purposes and does not replace advice from a doctor.