Childhood inflammatory and metabolic disease following exposure to antibiotics in pregnancy, antenatally, intrapartum and neonatally

Acute lymphoblastic leukaemia

In support of the hygiene hypothesis is recent evidence concerning the pathophysiology of acute lymphoblastic leukaemia. Acute lymphoblastic leukaemia is a disease of affluence and the commonest cause of childhood leukaemias. It is also the major cause of childhood cancer in high-income countries⁵. The incidence of acute lymphoblastic leukaemia is rising by 1% each year in high-income countries and is more common in first-born infants. The theory behind this finding is based on a two-stage gene–environment interaction. The first stage occurs in utero with an accidental gene mutation. The second stage occurs postnatally and makes the immune system more prone to react to infection, but only in children brought up in a relatively germ-free environment. According to Hawkes, “the first change loads the gun, the second pulls the trigger”⁵. In only-children raised in such a sanitised environment with little or no social contact in the first year of life, the development of the immune system is thought to be flawed. This flaw will be less evident in children who are breastfed and those born by vaginal birth. Accordingly, since the immune system evolves in an environment of infections, infection is required for it to form correctly⁶. Paradoxically, acute lymphoblastic leukaemia is caused by infections but also a lack of infection. The absence of “rough and tumble” in the first year of life leaves the immune system incapable of an appropriate response at a later stage. Although it could be argued that children today are exposed to infection through new vaccination programmes that stimulate the immune system, the antigens used to develop such vaccines are often based on attenuated viruses (as in varicella vaccine)⁷ or virulence factors such as capsular polysaccharides (as in group B haemolytic streptococcus vaccination)⁸ rather than the complete, wild organism that may be more antigenic and better able to stimulate an adequate immune response.

Conceptual considerations and evolutionary perspectives

The response to starvation selects for energy efficiency and favours the storage of excess calories when food is scarce. Conversely, with nutritional surplus, this once-advantageous metabolic state can lead to excess adiposity and its associated problems. The ability to resist infection has also led to Darwinian selection of a strong immune response. This is particularly evident when consideration is given to massive population declines during infectious epidemics and pandemics^11,12. This may be considered a form of “natural selection”. These traits in combination may generate an organism that is highly capable of processing and storing energy and is equipped with a powerful, and potentially hypersensitive, immune response. This relationship between the metabolic and immune response systems and the physiological mechanisms that control these responses in higher organisms has an evolutionary basis from common ancestry.

Evolutionary perspective. An example of this is the fat body of the fruit fly (Drosophila melanogaster), which incorporates the mammalian homologues of adipose tissue and the haematopoietic and immune systems¹³ and has similar functional and developmental pathways⁹. Drosophila’s fat body coordinates the appropriate metabolic status and pathogen survival responses. In contrast, in higher organisms, the liver and haematopoietic system and adipose tissue have specialised into discrete functional units or organs but have maintained their developmental heritage, shared with more primitive organisms. Accordingly, the most primitive response systems integrate the pathogen- and nutrient-sensing pathways such that nutrients can induce immune responses and, conversely, pathogens can evoke and regulate metabolic responses.

Taken together, inflammation and metabolic signalling create a delicate balance. Whereas the short-term compensatory and adaptive measures maintain this balance, when one arm overwhelms the other, the outcome is often detrimental in the long term. Accordingly, sustained exposure to pathogens can disrupt systemic metabolic function in both flies and humans. Similarly, chronic malnutrition or over-nutrition can disturb metabolic homeostasis and lead to an aberrant immune response. For those readers who wish to explore this theory further, linking inflammation with metabolic disease, we refer them to the excellent review by Hotamisligil⁹. For the purpose of our review, we simply want to establish an association between inflammation and metabolic disease which provides a theoretical link between obesity and alterations to the development of the immune system.

The relationship between antimicrobials given before, during or immediately after pregnancy and childhood obesity

In the past four decades, childhood overweight and obesity prevalence has risen substantially in most high-income countries¹⁴. Although the data are scarce, there also appears to be a rapid rise in childhood obesity in low-income and middle-income countries, despite continuing high levels of under-nutrition. Whereas data for children younger than 5 years have appeared in many large country surveys, data about older children and adolescents are less common, and sample sizes tend to be smaller. Nevertheless, comparable surveys show that the prevalence of overweight in these children is also rapidly increasing. A PubMed search indicates that, using the search term “childhood obesity prevention”, the number of published papers rose from about 20 per year in the late 1980s to 60 per year in the late 1990s and to more than 700 in 2018 alone¹⁵. At the time the manuscript of this review was submitted, there were already more than 600 such publications listed in 2019.

With an important role in adipogenesis, any alteration in the gut microbiome increases the susceptibility to obesity in later life¹⁶. A number of studies have addressed the use of antibiotics in pregnancy and subsequent childhood obesity^17–21, only one of which found no association between antenatal antibiotics and childhood obesity²¹. A large Danish study indicated that school-age children up to 16 years of age, exposed to antenatal antibiotics, had a nearly 30% rise in the prevalence of obesity¹⁷. In another study, antibiotics administered in the second or third trimester, resulted in an 84% greater risk of obesity by 7 years of age when compared with those children not exposed to antenatal antibiotics¹⁹. Delivery by caesarean section (CS) compared with vaginal birth was associated with a 46% greater risk of obesity regardless of whether by emergency or elective section¹⁹. In a cohort of 527 children, antibiotics (particularly those given in the first and second trimester), but not antifungals, were associated with a raised body mass index by 2 years of age²⁰. Although the weight of published evidence favours an association between the use of antibiotics in pregnancy and childhood obesity, we accept that obesity is a complex phenotype affected inter alia by maternal nutrition and health before and during pregnancy. There were also many confounding factors, not fully addressed in the studies cited, such as maternal obesity and diet during pregnancy, gestational diabetes, gestational weight gain and the number of courses of antibiotics administered.

Extent of use of antimicrobials in pregnancy

Although concerns over teratogenesis from the use of antimicrobials in pregnancy are not major, the use of thalidomide casts a long shadow, so we should not be complacent³⁴. However, antibiotic resistance remains a major problem³⁵. The extent to which antimicrobials in general and antibiotics in particular are used in pregnancy is difficult to quantify as these may be administered in hospital, prescribed by primary care physicians in the community, or self-administered by pregnant women in the form of antifungals or antivirals purchased over the counter to treat self-diagnosed conditions such as herpes labialis, herpes genitalis or vulvovaginal candidiasis. Antimicrobials may be used for infections such as urinary tract infections, which are more common in pregnancy because of the physiological adaptations that are due to progestagenic effects on smooth muscle. Alternatively, they may be used for prophylaxis following preterm prelabour rupture of the membranes or to prevent early-onset neonatal group B streptococcal infection in women at risk of passing this on to the neonate or at the time of CS. A conservative estimate of 40% can be made from a Danish Registry–based study of nearly a million women³⁶ but that is likely to be an underestimate as it applied only to antimicrobial prescriptions filled in the community without data on hospital-based or self-administered antimicrobials.

In a recently published study of 5.6 million pregnancies in the US between 2006 and 2015, 57% of women received antibiotics during hospitalisation for delivery³⁷. The focus of the study was the risk of Clostridium difficile infection associated with the use of clindamycin and acute kidney injury associated with gentamicin during hospitalised birth. Although the absolute risk of such complications was low, the receipt of clindamycin was associated with significantly increased likelihood for C. difficile infection and receipt of gentamicin with significantly increased likelihood of acute kidney injury. Nevertheless, the impression given was that there was pressure for an increased, not decreased, use of antibiotics during hospitalisation³⁷. This dichotomy may be emphasised by the recommendations for the early use of antibiotics when sepsis is suspected³⁸.

Allergy and atopic disease

In 1988, following successive reports from British General Practice Surveys, hay fever was described as a “post–Industrial Revolution epidemic”³⁹. The following year, it was noted that the prevalence of asthma and childhood eczema (allergic dermatitis) had been increasing over the previous 30 years. Currently, atopic disease, a syndrome of hypersensitivity which typically presents as one or more of an “allergic triad” of asthma, eczema and hay fever, affects around 25% of the population, particularly in industrialised countries and especially in children and young adults⁴⁰. The use of antenatal or intrapartum antibiotics is significantly associated with atopic dermatitis (eczema) in infancy if they are used for more than 24 hours⁴¹ or if the infant was delivered by CS⁴². In the first year of life, cow’s milk allergy (lactose intolerance) accounts for up to 6% of food allergies⁴³. Food intolerance is probably due to a mixture of genetic and environmental factors but may be due to modifiable risk factors such as the use of antibiotics in pregnancy. Two studies reported that the use of antibiotics before or during pregnancy led to an increased risk of lactose intolerance in offspring, although this was less in children who were breastfed for longer periods of time^44,45.

Asthma

Over the last four decades, the global incidence of asthma has increased; it is estimated that there were 334 million cases in 2014^46,47. Asthma is also the most common chronic disease in children between 5 and 14 years of age⁴⁶. Among the risk factors for asthma in offspring are maternal asthma, exposure to cigarette smoke, and air pollution in pregnancy^48,49, together with low birth weight and preterm birth⁵⁰. Even after controlled for confounding factors, an increasing number of studies have concluded that there is a positive association between childhood asthma and the use of antibiotics in or around pregnancy^51–56. Vitamin D (the “antibiotic vitamin”) is an immunomodulator that is associated with a number of outcomes in pregnancy⁵⁷ and extra-skeletal outcomes in children⁵⁸. Supplementation with vitamin D in pregnancy reduces the risk of infection and the need to use antibiotics in pregnancy and also reduces recurrent wheeze and asthma in offspring in childhood at the age of three years⁵⁹ though not at the age of six. However, care should be taken while interpreting such data. Two large registry-based studies^55,56 suggested that the association between maternal antibiotic use and asthma might be due to a maternal propensity for infection rather than the use of antibiotics per se. The association was also stronger when antibiotics were prescribed for respiratory infections. A large population-based cohort study of mother–child pairs⁶⁰ supported the contention that the association between maternal antibiotic use and asthma was not unique to use during pregnancy. However, it did provide evidence that maternal antibiotic exposure in a dose-dependent manner was associated with an increase in the risk of childhood asthma. After adjusting for infant gender, mode of delivery and method of infant feeding, most classes of antibiotics were associated with an increased risk of asthma, although the degree of risk varied.

The Neomune Project

The Neomune Project was developed by the University of Copenhagen to study the role of diet and early microbial colonisation in the development of the neonatal immune system and cognitive function in vulnerable preterm or small-for-gestational-age infants. One aim of the Neomune Project is to investigate dietary and microbial elements that help maturation of the immune system in order to develop formula milk to support infants for whom breastfeeding cannot occur or is contraindicated. Another aim is to supplement the limited research on the use of probiotics in milk products. It is thought that these probiotic supplements have the potential to support the neonatal gut microbiome and to improve development of the immune system while suppressing pathogens in vulnerable neonates. For more information, see https://food.ku.dk/english/research_at_food/research-projects/2014/neomune/.

Seeding therapy

First proposed in 2016, vaginal seeding (sometimes referred to as microbirthing) involves the postnatal transfer of maternal vaginal fluid to neonates born by CS, to correct the disruption of the neonatal microbiota that occurs with CS compared with vaginal birth⁶¹. The evidence to provide confidence that vaginal seeding is safe in humans is not robust⁶², and there is a risk of neonatal transfer of asymptomatic vaginal pathogens such as herpes simplex virus, Neisseria gonorrhoeae and Chlamydia trachomatis. Since intrapartum antibiotic chemoprophylaxis for the prevention of early-onset group B streptococcal infection of the neonate in the UK and many other European countries is risk-based rather than screening-based⁸ (as occurs in the USA), there are concerns that this significant cause of neonatal sepsis could be inadvertently transferred to the neonate by vaginal seeding. In the UK and Denmark, obstetricians advise against vaginal seeding and are required to inform women of the potential risks if they intend to pursue this route independently. The introduction of probiotics and breastfeeding are promoted by neonatologists. Ideally, new information about the species-specific function of Lactobacilli from the vaginal microbiome project⁶³ will yield additional data for the development of improved probiotics^64,65.

Breastfeeding

As the first type of food that is introduced into the infant gut neonatally, breast milk is thought to modify the developing microbiota, and, in babies breastfed within the first hour of life, there are similarities between the bacterial constituents of maternal colostrum and new-born meconium⁶⁶. Species of Bifidobacteria are dominant in human breast milk, protect the infant gut from pathogens, and provide health benefits later in life. However, numbers of these bacteria are reduced in the breast milk of women who deliver preterm and in those born by CS, although whether this relates to prophylactic antibiotic cover is unclear⁶⁷. Accordingly, the American Academy of Pediatrics and the Royal College of Paediatrics and Child Health highlight the importance of breastfeeding and advocate the use of donor breast milk where maternal feeding is contraindicated, or not possible, rather than resorting to the use of formula feeds⁶⁸.

Probiotics

The World Health Organization defines probiotics as “live organisms, which when administered in adequate amounts, confer a health benefit on the host”⁶⁹. Although probiotics purport to balance the gut flora, reliable information is lacking with respect to materno-fetal and neonatal safety. In neonatal mice, T-cell disturbance secondary to the use of antibiotics was prevented by probiotics⁷⁰, and improved mucosal tolerance following probiotics was reported⁷¹. Probiotics may reduce the incidence of pulmonary tuberculosis as well as allergies and asthma in children. However, a higher level of Bifidobacteria in infants given probiotics lasted only one week and was associated with an increased risk of infections in later life⁷². Accordingly, further well-conducted research is required; this will be covered as part of a separate review. Although increased diversity of the gut microbiota is considered a beneficial attribute, the opposite is true of the vaginal microbiota. New information about the vaginal microbiome is providing data about species-specific function of eubiotic, lactic acid–producing organisms that may lead to better probiotic therapies⁶³, but the microbiome of the gut and the vagina are sufficiently complex that the use of a single probiotic, or even two probiotics, may not be able to address this complexity.