Introduction

Burden of Otitis Media

Acute otitis media (AOM) is usually caused by either a viral or bacterial pathogen and often treated with antibiotics. Young children are particularly at risk due to limited space in the middle ear and poor drainage from relatively tortuous Eustachian tubes. AOM is often accompanied by significant pain along with fever, acute loss of hearing, and general unwellness. AOM has a number of potential complications. Many AOM-prone children suffer from recurrent episodes in infancy and early childhood. Commonly, on resolution of the acute infection, there is a persistent middle ear effusion (otitis media with effusion-OME) with accompanying hearing impairment. OME can persist, causing longer term hearing impairment and interfering with language development, school performance, and behavior. AOM may result in perforation of the eardrum with chronic discharge (chronic suppurative otitis media). Less common complications include mastoiditis, brain abscess, and meningitis.

Globally, there are an estimated 709 million cases of acute otitis media per year, an incidence of 10.85 % [1]. The peak incidence is in the 1–4-year age group (60.99 %), although rates vary from 3.64 % in Central Europe to 43.36 % in Sub-Saharan Africa [1]. The disease burden attributed to AOM in both established and emerging nations is considerable. It is the most common reason for prescription antibiotics in developed countries where AOM is estimated to affect more than 60 % of children under 1 year and more than 80 % of children under 3 years [2–4]. There is also a considerable health burden from the sequelae of AOM. It is estimated that there are globally around 31 million cases of chronic suppurative otitis media, and the prevalence rate for AOM induced permanent hearing impairment is 30.82 per 10,000 [1]. Furthermore, complications of AOM are estimated to cause the deaths of 21,000 people annually, with mortality rates being the highest in the 0–5-year age group [1].

Although updated clinical guidelines have reduced the reliance on antibiotic therapy for mild, early and uncomplicated AOM [5], it is still the most common reason for antibiotic use in many countries [6]. Apart from the associated health care burden, this widespread use of antibiotics increases the risk of community antibiotic resistance.

Given the high health care burden, a number of risk factors for AOM have been investigated. The factors currently identified for risk reduction are breastfeeding, avoidance of in utero and childhood passive tobacco smoke exposure, and avoidance of indoor air pollution [1]. A recent Lancet mega-review that summarized the evidence on breastfeeding and maternal and child health outcomes from 28 individual systematic reviews found that breastfeeding has many health benefits for both mothers and children [7••]. The Lancet findings also confirm the vital role that breastfeeding plays in the prevention of common childhood infectious diseases including AOM.

Breastfeeding and Otitis Media

Human breast milk is tailor-made for infants [8]. Breast milk delivers appropriate nutrition for each infantile developmental stage and is packed with immune substances that may directly influence microbial colonization with favorable bacteria, protect against colonization and infection by harmful bacteria, and influence immune programming [9]. The WHO recommends exclusive breastfeeding up to the age of 6 months with continued breastfeeding for 2 years and beyond. Despite these recommendations, breastfeeding rates in many countries, especially in high-income countries, are poor, with rates of only around 20 % at 12 months [7••]. Additionally, in children under the age of 6 months, 63, 61, and 55 % are not exclusively breastfed in upper-middle-income countries, low-middle-income countries, and low-income countries, respectively [7••].

Almost all the recent literature outlining the links between breastfeeding and OM supports a reduced risk of OM associated with breastfeeding. Since 2013, there have been two systematic reviews, a non-systematic review and four original studies on this subject.

Data Syntheses: Systematic and Non-systematic Reviews (Table 1)

Table 1 Reviews of breastfeeding and otitis media in the past 4 years
Full size table

There have been three reviews published on the impact of BF on OM over the last 4 years.

The most recent data synthesis on this topic is our group’s 2015 systematic review and meta-analysis by Bowatte et al. [10], commissioned by the World Health Organization as part of the mega-review recently published in the Lancet [7••]. Pubmed, Cinahl, and Embase databases were searched from inception yielding 24 studies (18 cohorts and 6 cross-sectional). Overall, we found a 43 % reduction of the risk of ever having AOM in the first 2 years of life associated with breastfeeding but no reduced risk after the age of 2. In this systematic review, meta-analyses were possible only for particular exposure categories. Infants exclusively breastfed for 6 months compared with those not breastfed or breastfed for periods less than 6 months had a reduced risk of AOM up to the age of 2 years after pooling the ORs of five cohort studies (OR 0.57; 95 % CI 0.44–0.75). We also found a risk reduction when comparing ever versus never breastfed infants: OR 0.67; 0.56–0.80 (five studies). An additional meta-analysis was performed on 12 cohort studies in a more versus less exposure grouping. This category included ORs from all studies which compared a greater exposure of breastfeeding (more) to less breastfeeding. The meta-analysis also found a reduced risk of AOM up to 2 years: OR 0.76; 0.67–0.56.

A systematic review by Hornell et al. reviewed the literature published between January 2000 and June 2011 [11]. They identified four publications on AOM; two systematic review/meta-analysis [12, 13] and two prospective cohorts [14, 15]. After reviewing these publications without performing an overall pooled estimate, the authors concluded that there was convincing evidence of a protective dose and duration of breastfeeding on OM. The publications included in the Hornell et al. systematic review are described in the following sentences. The included systematic review by Ip et al. [12] found a pooled adjusted odds ratio from five cohort studies for the risk of AOM associated with any breastfeeding of 0.77; 95 % CI 0.64–0.91 when compared with never breastfed infants. Additionally, Ip et al. found some evidence that longer duration of breastfeeding may confer greater protection; the pooled estimate for the risk of AOM associated with 3–6 months exclusive breastfeeding versus never breastfed was OR 0.5; 0.36, 0.70. The other included systematic review by Kramer and Kakuma [13] analyzed two prospective cohorts with a total of 3762 children finding an increased risk of one or more episodes of otitis media in the first 12 months for children exclusively breastfed for more than 6 months compared with those who had exclusive breastfeeding for 3 months (risk ratio 1.28;95 % CI 1.04–1.57). One of the two additional cohort studies included in the Hornell et al. systematic review (birth cohort n = 1764) found a non-significant association between distinct lengths of breastfeeding exposure (<1, 1–3, 4+, 4–6, 7–11, and 12+ months) and prevalence of ear infections in either the first or second 6 months of life. Their findings may have been limited by lack of power in each of the exposure categories. The remaining cohort study on 926 children found that infants exclusively breastfed for 6 months had fewer infections than partially breastfed or never breastfed children (OR 0.37; 0.13, 1.05).

The Bowatte review included all studies within the Hornell review and identified an additional 15 studies related to the timing of the search (Bowatte search conducted in 2014 versus 2011) and the search inclusion criteria ( Hornell limited to studies published after 2011 and Bowatte had no limits).

A 2013 non-systematic review performed by the American Academy of Pediatric Dentistry, Chicago as an update on the effects of breastfeeding for dental professionals also found that breastfeeding was associated with a reduced risk of OM [16].

Original Articles since 2013 (Table 2)

Table 2 Original studies on the link between breastfeeding and otitis media—published in the last 4 years
Full size table

There have been four original research articles published since 2013 [17–20]. Due to dates of publication (two studies [17, 19]) and inclusion/exclusion criteria (two studies on selected populations) [18, 20], none were included in either of the systematic reviews mentioned above; however, it is unlikely that their inclusion would have affected the direction of associations found as three of these articles found an association between breastfeeding and a reduced risk of AOM.

Most recently, Martines et al. [19] performed a case–control study on Sicilian children (204 cases with 204 age and sex matched controls). They found that children who were breastfed were much less likely to develop AOM or OME following an upper respiratory tract infection (URTI) than those who had never been breastfed; OR 0.5;95 % CI 0.3–0.77 [19].

Ajetunmobi et al. [17] investigated 502,958 children in a retrospective population-based Scottish cohort using linkages of birth, death, maternity, infant health, child health surveillance, and admission records. They included all single births in Scotland between 1997 and 2009, following the children until March 2012. Based on information collected about feeding at the 6–8-week visit, infants were classified as either exclusively breastfed, exclusively formula fed, or mixed fed. Compared with exclusively breastfed children, there was an increased risk of hospitalization for AOM in the first 6 months of life for infants who were exclusively formula fed (hazard ratio (HR) 2.13; 95 % CI 1.26–3.59). This estimate was made following adjustment for a range of socio-economic factors. There was also an increased point estimate for those both breast and formula fed (mixed feeding) compared with exclusive breastfeeding but the 95 % confidence interval included 1 (HR 1.5; 0.65–3.48). They also found increased risk of hospitalization within the first year of life for formula-fed infants for a large range of illness including infections (gastrointestinal, upper and lower respiratory tract, urinary, and non-specific fever), asthma, diabetes, and dental caries. These increased risks persisted after stratification by area deprivation. There was no increased risk of hospitalization for AOM after the first 6 months of life. The lack of association after the age of 6 months is perhaps expected considering that AOM is usually treated in the community, and it is also remarkable that an association was found up to the age of 6 months. The huge population-based sample, along with the objective ascertainment of prospectively collected exposure and outcome data, contributes to the robustness and importance of this work.

Jensen et al. [18] investigated a population-based cohort of 223 Inuit mother-child pairs in two towns on the west coast of Greenland (1999–2007). The primary purpose of this research was to assess the relationship between maternal organochloride exposure and OM in their children. The children were followed up at the age of 4–10 years. They classified breastfeeding status at 6 months as full, partial, or not and did not find an association with breastfeeding although point estimates were below 1 suggesting protection for full or partially breastfed children. It may be that there was little power to detect an association given the number of participants.

The final article by Salah et al. was a retrospective hospital-based cohort of 340 children. In a group of children aged less than 2 years attending an outpatient clinic for recurrent AOM (three or more episodes in 6 months), factors were analyzed which predicted further recurrence and treatment failure. They found that breastfeeding duration of less than 3 months (compared with more than 3 months) was associated with a significant chance of further recurrence and with treatment failure (failure of antibiotic treatment).

Mechanisms for the Protective Link Between Breastfeeding and AOM

Previously, it was believed that the protective effect of AOM on breastfeeding was largely mechanical; the suction pressure required and positioning for breast feeding were thought to be advantageous for draining the Eustachian tubes in young infants, thereby preventing AOM. Another slightly older theory that may underlie the reduction in infectious disease enjoyed by breastfed babies is related to the immunomodulatory substances contained in breast milk. This theory has currently been re-invigorated through the recent interest in the human and more specifically the infant gut microbiome. It is now believed that a specific symbiotic microbiome is established early in life and, among other functions, protects the infant against pathogenic infections [21].

Breast milk is known to contain the building blocks for establishment of this microbiome in the form of human milk oligosaccharides (HMOs) along with a distinct breast milk microbiota. HMOs are indigestible sugars most prominent in the colostrum. The milk oligosaccharides from primates are unique in the mammalian kingdom in terms of their diversity and high percentage of fucosylation [22]. There are over 200 different types of human HMOs [23•], and 50–80 % of these are fucosylayted depending on the genetic makeup of the mother [24]. They feature prominently in the colostrum (20–25 g/L) and taper off in overall percentage for mature breast milk (5–20 g/L) [25]. Although these HMOs do not provide a source of energy for the infant, their unique branching and diversity make them a perfect substrate for particular strains of bacteria that are known to be beneficial colonizers of the newborn.

Although there is a growing literature on the establishment of the infant gut microbiome and its importance for protection from gastrointestinal morbidity along with its capacity to correctly educate the infants’ immature immune system, less is known about the microbiome of the nasopharynx and its potential effect on protection from AOM. Recently, Biesbroek et al. [26••] investigated the nasopharyngeal microbiome in 101 exclusively breastfed and 101 exclusively formula-fed infants. They discovered a distinctly different bacterial community composition in the nasopharynx between the two feeding modes; with breastfed children having increased representation of Dolosigranulum and Corynebacterium Sp. and reduced representation of Staphylococcus, Prevotella, and Veillonella spp. at 6 weeks of age.

Literature Concerning Cost Savings from Reduced AOM Linked to Increased Breastfeeding

Given the convincingly positive findings for breastfeeding in relation to OM, some of the recent literature has focused on reductions in OM and cost savings which could be achieved through greater uptake and continuation of breastfeeding. Pokhrel et al. [27] assessed the potential economic impact, from the point of view of the National Health Service, from improving breastfeeding rates in the UK which are comparatively low internationally with only 55 % of infants breastfed at 6 weeks and only 23 % exclusively (2010). They assessed the economic impact in terms of four acute childhood conditions in the first year of life: gastrointestinal illness, lower respiratory tract infections, AOM, and necrotizing enterocolitis, finding that for women who have initiated breastfeeding for the first week, an increase in breastfeeding duration up to 4 months would save 11 million pounds per year. For AOM, these costs were based solely on the costs of treating AOM in primary care and did not appear to take into account the cost of complications, hospitalizations, or the costs for parents who would need time off work to care for their children. Specifically, for AOM, the cost saving was estimated to be between 0.28 and 1.16 million pounds per year depending on whether exclusive breastfeeding rates at 6 months increased to 21 or 65 %, respectively. McIsaac et al. [28] studied potential reductions in common childhood infections in Aboriginal Canadians where AOM reduction may be arguably greater due to the increased prevalence of severe AOM in this population and relative decrease in breastfeeding when compared with general Canadian infants. They found a 5.1 to 10.6 % reduction in OM in Aboriginal infants if they received any breastfeeding. The preventable proportion of infectious disease in Aboriginal infants was 1.5–2 times greater than the non-Aboriginal Canadian infants. Arantxa Colchero et al. [29] investigated the costs of inadequate breastfeeding in Mexican infants who experience very low rates of exclusive breastfeeding at 6 months; only 14 % in 2012. They found that if exclusive breastfeeding rates increased to 95 % at 6 months and 95 % partial breastfeeding between 6 and 12 months, then the savings related to reduced AOM could be between US $0.5 and 15.4 million per year. This estimate increased when the cost of infant formula was added: US $289.9 million dollars per year. Furthermore, the economic modeling from the Lancet mega-review using the Lives Saved Tool to estimate global impacts of increasing exclusive breastfeeding to 95 % of all children at 1 month and 90 % at 6 months with partial breastfeeding of 90 % between 6 and 23 months found a possible prevention of 823,000 deaths in children under the age of 5. This was largely from prevention of infectious disease in low-income countries [7••].

Conclusions

There appears to be little doubt that breastfeeding is beneficial for protecting infants from a range of infectious diseases including AOM. In addition, there are a number of other economic and health-related reasons for promotion of breastfeeding in line with the WHO guidelines. Governments globally should promote and support breastfeeding through campaigns and measures designed to educate and support mothers, families, and communities.