Role of probiotics to prevent and reduce the duration of upper respiratory infections in ambulatory children : Systematic Review with Network-Meta Analysis

Background. Upper respiratory infections (URIs) remains as significant cause of morbidity in children. Evidence on efficacy of probiotics to prevent URIs in children is increasing. This systematic review was assembled to analyze evidence about the efficacy of probiotics to reduce duration of upper respiratory infections in ambulatory children. Methods. Randomized controlled trials (RCTs) comparing probiotics vs. placebo to prevent URIs, published between 2001 and 2016 were considered. Quality evaluation was evaluated using CONSORT. Standard mean difference (SMD) or risk ratio (RR) was calculated. Network Meta-Analysis (NMA), using a random effect model was assembled. Results. 31 RCTs were evaluated and 20 studies were included with 3,635 children randomized to probiotics and Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 1 October 2018 doi:10.20944/preprints201810.0002.v1 © 2018 by the author(s). Distributed under a Creative Commons CC BY license. 3,433 to placebo. Lactobacillus reuteri [SMD -0.56 CI95% (-0.72 to -0.41), p 0.0001] and Lactobacillus acidophillus [SMD -0.33 CI95% (-0.60 to -0.06), p 0.01] were superior to placebo to reduce duration of URIs. L. rhamnosus GG showed tendency [SMD -0.14 CI95% (-0.28 to 0.0), p 0.048]. On the network forest plot L. reuteri showed preventive equivalence when was compared to L. rhamnosus GG, L. casei and BB12. Conclusions. Lactobacillus reuteri, Lactobacillus rhamnosus GG and Bifidobacterium BB12 are evidence-based alternatives to be considered to prevent URIs in children.


Introduction
URIs are a common problem in the first decade of life.The yearly prevalence of respiratory tract infections in an otherwise healthy 3-year old child is about three to 10 infections.(1)(2)(3)(4) Although most respiratory tract infections are generally self-limiting and in more than 70% are from viral etiology, the use of antibiotics around the world is very high.Over a time period of 12 months, between 30% to 75% of all children and adolescents younger than 15 years of age takes antibiotics for URIs.(5) About 85-88 % of all respiratory infections in children are URIs with pharyngitis/tonsillitis as the most frequent URIs.(6)(7)(8) This type of infections imposes a significant economic burden on health systems and individual families in developing countries.We recently estimated that among children aged < 5 years, the median direct cost of ARI was US$135 in private and US$54 in public institutions.(9) Considering that in the most of cases URIs in ambulatory children are from viral etiology, that there is no consensus to start treatment of URIs with antibiotics, taking into account all the antibiotic associated adverse events (10), and secondary to the effects on immunity and inflammation previously proved for some probiotics, there has been a growing interest in the use of this alternatives for treatment and prevention of this type of infections in children and adults.(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22) The first randomized controlled trial (RCT) about the use of probiotics to prevent URIs in outpatient children was published in 2001.It was conducted in 18 daycare centers in Helsinki, Finland.571 healthy children aged 1 to 6 years were included (n=282 in probiotic group and 289 in control group).Lactobacillus rhamnosus GG was administrated in milk.Children in the Lactobacillus group had fewer days of absence from day care because of illness [4.9 (95% confidence interval 4.4 to 5.5) vs 5.8 (5.3 to 6.4) days, 16% difference, P = 0.03.There was also a relative reduction of 17% in the number of children suffering from respiratory infections with complications and a 19% relative reduction in antibiotic treatments for respiratory infection [unadjusted absolute % reduction -9.6 (-18.2 to -1.0), P = 0.03; adjusted odds ratio 0.72 (0.50 to 1.03), P = 0.08] in the Lactobacillus group.(11) Twelve years later the first meta-analysis was published.Four RCTs involving 1805 participants were considered.Authors concluded that compared with placebo, probiotics administration was associated with a reduced incidence of acute otitis media (four RCTs, n=1805, RR 0.76, 95% CI 0.64-0.91,fixed effects model, NNT 17,, a reduced risk of URIs (one RCT, n=281, RR 0.62, 95% CI 0.50-0.78,NNT 4, 95% CI 3-8) and antibiotic treatments (four RCTs, n=1805, RR 0.80, 95% CI 0.71-0.91,fixed effects model).Unfortunately, the number of RCTs that supported the reduction of risk for URIs was only one, establishing the need to assembly other RCTs.(12) In 2014 and 2015 were published others meta-analysis on the same topic, including 10 and 12 RCTs respectively.(13,14) Unfortunately, the included studies in children and adults and with specific populations (i.e.malnourished patients), what made recommendations difficult to extrapolate to healthy children.In 2016, a fourth meta-analysis was published, which included only RCTs performed in children.A total of 23 trials involving 6269 children were eligible.None of the trials showed a high risk of bias.The quality of the evidence of outcomes was moderate.The age range of subjects was from newborn to 18 years.The results of meta-analysis showed that probiotic consumption significantly decreased the number of subjects having at least one URI episode (17 RCTs, 4513 children, relative risk 0.89, 95% CI 0.82-0.96,P=0.004).Children supplemented with probiotics had fewer numbers of days of RTIs per person compared with children who had taken a placebo (6 RCTs, 2067 children, MD 0.16, 95% CI 0.29 to 0.02, P=0.03), and had fewer numbers of days absent from day care/school (8 RCTs, 1499 children, MD 0.94, 95% CI 1.72 to 0.15, P=0.02).However, there was no statistically significant difference of illness episode duration between probiotic intervention group and placebo group (9 RCTs, 2817 children, MD 0.60, 95% CI 1.49 to 0.30, P=0.19).(15) Unfortunately, this meta-analysis included not only RCTs focused on prevention, but also RCTs assembled for treatment of respiratory infections.Finally, a Network Meta-Analysis (NMA) was published recently.It was focused on the effects of probiotics to prevent respiratory infections in children and adolescents.Twenty-one trials with 6.603 participants were included.Pairwise meta-analysis suggested that Lactobacillus casei rhamnosus (LCA) was the only effective probiotic to the rate of RTIs compared to placebo (RR0.38;CI 0.19-0.45).NMA showed that the LCA exhibited 54.7% probability of being classified in first, while the probability of Lactobacillus fermentum CECT5716 (LFC) being last in the ranking was 15.3%.LCA showed no better effect compared to other probiotic strains by indirect analysis.(16) When this paper was reviewed by our group we found some papers assembled with different objectives (i.e. to reduce the risk of necrotizing enterocolitis and/or neonatal sepsis in preterm (17); to reduce the cases of recurrent otitis media (18,19); to prevent nosocomial infections (20,21) or to reduce the presence of bocaviruses in mouth (22)].Considering limitations of previous reviews, we decide to assembly this new systematic review with NMA limited to evaluate the strain specificity of Lactobacillus reuteri and other probiotics to prevent and reduce the occurrence and duration of upper respiratory infections in ambulatory children.

Materials and Methods
Systematic review and network meta-analysis was assembled according to PRISMA statements (23), incorporating Network Meta-analyses of Health Care Interventions (24), and approved by the Research Committee at Hospital General Dr. Manuel Gea Gonzalez, Ministry of Health.Mexico.(Supplementary materials)

Search strategy
We included in this systematic review only double-blind, randomized, controlled clinical trials (RTCs), published between January 2001 and January 2016, in English or Spanish language.An exhaustive search was conducted in Medline, Embase, Cumulative Index to Nursing and Allied Health (CINAHL), PsycINFO, the Cochrane Central Register of Controlled Trials, Lilacs, Artemisa and in the databases of the principal international regulatory agencies in order to identify relevant studies published.

Study selection and outcome measures
RCTs that compared the use of probiotics (mono-strain or combined strains) versus placebo or active treatment in ambulatory or attending daycare centers infant (birth to 23 months of age), preschool child (2-5 years old) or child (6 to 12 years old) were consider to be included in this network meta-analysis.All different probiotics used to prevent and/or reduce the occurrence of URIs were included in 14 nodes of intervention (Figure 1).Primary outcome analyzed was the reduction of the duration of the episodes of URIs observed during intervention.Secondary outcomes were the reduction of a) at least one episode of URIs; b) the number of days with URIs; c) the use of antibiotics and d) the days of scholar absentees.Data extraction and quality analysis Quality evaluation of studies was performed in pairs, in a blinded and independent fashion using CONSORT statement for RCTs.(25) Any discrepancy in the evaluation of the articles was resolved using Delphi methodology, which was coordinated by the principal investigator.Analysis included characteristics of participants (age, country, type of URI), types of intervention (strain of probiotic, CFU dose administrated, time of administration) and reported outcomes (duration of the episodes, number of episodes, number of days with URI, use of antibiotics and scholar absentees) Risk of Bias (RoB) in the Included Studies Trial quality and RoB in the included studies were assessed according to the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (26), taking into account evidence interpretation using the GRADE approach.This included rating the adequacy of randomization: randomization; allocation concealment; blinding participants; blinding outcomes; incomplete outcome data; selective reporting; vested interest bias and loss of participant follow-up.

Data synthesis and analysis
From a statistic point of view, the information was analyzed with the strategy of multiple treatments meta-analysis.We use placebo groups as the central axis for direct comparisons.Indirect comparison was one probiotic against other.Dichotomous outcomes were analyzed with the total number of randomly assigned participants as the denominator.When reported, information on participants that abandoned the studies was included in the analysis.For each potentially eligible study, descriptive statistics of the population characteristics and their results were reported, describing the type of comparison as well as the most important clinical and methodological variables.For each pairwise comparison (direct or indirect), the Hedges standard mean difference (SMD) was calculated for continuous numeric variables, whereas the respective risk ratio (RR) was calculated for dichotomous outcomes.Both were calculated with their respective 95% confidence interval (CI95%).The first meta-analysis were paired comparisons of all published studies, stratified for strain.For this analysis, we used a random effect model, considering that different studies estimated different treatment effects.Concomitantly, we calculated I 2 for heterogeneity and its corresponding P value.Thereafter we assembled a NMA, using a random effect model with a Bayesian approach (27,28) and summarized the results using effect sizes and CI95%.We used the adjusted model as described by Salanti et al. ( 29) Additionally, we calculated the probability of superiority for each "strain of probiotic" intervention through a SUCRA analysis and presented the results in a ranked graph.(30) To estimate the inconsistency (discordance between direct and indirect evidence with a CI95% that did not include zero), we calculated the difference between the direct and indirect estimates, taking as reference only the constructed indicators that had included a placebo.( 31) Finally, we adjusted the model with and without assumptions of consistency and compared the two models in terms of fit and parsimony.(32) In the case of a significant inconsistency we investigated the distribution of clinical and methodological variables that might have been a potential source of heterogeneity or inconsistency in each group of specific comparisons.All analysis and graphic depictions were performed on the version 15.1 of STATA for Mac

Discussion
URIs are one of the most common reasons for children in the USA to seek medical care (63), resulting in billions of dollars in health care expenditures; 30 % of URIs are complicated by acute otitis media and 8 % by rhinosinusitis.Considering that in the most of cases the etiology is viral, there had been a lot of debate regarding the early use of antibiotics.Unnecessary use of antibiotics to treat such conditions poses significant financial burden and can result in untoward side effects as well as risk of antimicrobial resistance.On the other hand, inadequate antibiotic treatment in certain cases may increase the risk of suppurative complications and/or invasive infection in this population.(64) Recently a prospective observational study was conducted in order to compare the treatment outcomes between those with and without antibiotic treatment for the uncomplicated URIs and acute diarrhea in young children.200 previously healthy children presenting with acute uncomplicated URI (aged 2 to 5 years) and 199 children with acute diarrhea (aged 6 months to 5 years) were included.The decision for antibiotic prescription was based entirely on attending physicians' discretion.Antibiotic prescription rates for URI and diarrhea groups were 30.2% and 13.6%, respectively.Univariate analyses indicated that odds ratios (OR) for treatment failure comparing those with and without antibiotics were 0.5 (CI95% 0.2 to 1.7) and 1.5 (CI95% 0.6 to 3.7) for URI and diarrhea, respectively.Logistic regression analyses indicated that antibiotic treatment was not significantly associated with better treatment outcomes neither URI nor diarrhea cases.Adjusted OR for antibiotic for requirement of additional treatment were 1.06 (CI95% 0.14 to 8.15) for URI cases.(65) On the other hand, some papers recently published had pointed out the importance of changes in the nasopharyngeal microbiome as significantly associated with the development of URIs.In 2016, the nasopharyngeal microbiome was analyzed in a cohort of 47 healthy children from USA, aged 49 to 84 months trying to establish a relationship between specific profiles and the developing of URIs.All NP samples were profiled for bacterial microbiota using a phylogenetic microarray, and these data were related to demographic characteristics and upper respiratory health outcomes.(11, Different to previous papers (13-15) our systematic review with NMA establish a strain-specificity analysis to clearly demonstrate the concept that in this moment is prevalent around the world in terms of strain-specificity, dosespecificity and frequency-specificity for a specific type of health/disease condition, showing the superiority of some strains of probiotics (L.reuteri, L. rhamnosus GG and Bifidobacterium BB12), for reduce the burden of the URIs in children.One of the results identified in our study, that is very important to emphasize, is the observed reduction for the use of antibiotics.Since the emergence of antibiotic resistance and the increasing frequency of antibiotic associated adverse, indiscriminate use of antibiotics has emerged as a relevant problem around the world.The prevalence of gastrointestinal, hepatic, renal and even hematological events observed during and after the use of antibiotics represents a health problem associated with very high costs (10), therefore identifying prevention alternatives that help to reduce the use of antibiotics represents a relevant public health strategy.In terms of mode of action and trying to establish a rationale about observed clinical effects, in contrast to how microbiota protects against intestinal infection, where a diversity of mechanisms have been clearly described for how the microbiota inhibits the establishment of colonization by enteric pathogens and fortifies innate and adaptive defenses against those that are able to invade this site (68,69), the mechanistic basis for how the microbiota protects against infection outside the intestine had been limited explained.Microbiota and different probiotics has been shown to protect against systemic and respiratory infection by Escherichia coli (70,71), influenza virus (72) Klebsiella pneumoniae (73,74) Listeria monocytogenes (75), Staphylococcus aureus (76), and Streptococcus pneumoniae (77)(78)(79).Very recent research had demonstrated that microbiota and probiotics protects against infection by a) diverse respiratory pathogens; b) enhances respiratory immunity via granulocyte-macrophage colony-stimulating factor (GM-CSF) signaling; c) IL-17A modulated for probiotics primes GM-CSF during infection downstream of the microbiota; d) regulates alveolar macrophage function via GM-CSF; e) GM-CSF stimulates extracellular signal regulated kinase (ERK) to promote bacterial killing by alveolar macrophages; f) modulate pattern recognition receptor (PPR) ligands; and g) stimulate Nod2 to regulate respiratory immunity.Activation of NOD2 seems to represent one of the most powerful mechanisms to protect against respiratory infections.In this sense, research has identified potent activators of NOD2 (activation of 36 ± 1.9), among which are L. reuteri, L. crispatus and Clostridium orbiscindens and weak activators (activation of 2.6 ± 0.76) such as Lactobacillus johnsonii and Lactobacillus.Rhamnosus.(80)

Conclusions
Probiotics are safe alternatives to reduce the frequency and duration of URIs in children.
They additionally help to reduce the use of antibiotics and the days for absentees.
In terms of strain-specificity there are preventive equivalence for L. reuteri, L. rhamnosus and Bifidobacterium BB2 for reduce the aforementioned outcomes.
Considering the economic impact and burden disease of antibiotic associated side effects, the associated reduction of antibiotics with the preventive use of probiotics for URI in children must be considered as public health strategy.

Fig. 2 Fig. 8 Fig. 10 Best
Fig. 2 Network Meta-Analysis on the efficacy of probiotics to reduce the ocurrence of upper respiratory infections in ambulatory children

Table 1 .
Included studies

Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 1 October 2018 doi:10.20944/preprints201810.0002.v1
(67)ors found that time to develop URI was positively correlated with NP diversity, and children who experienced more frequent URIs exhibited significantly diminished NP microbiota diversity, p <0.05.(66)Very recently 971 specimens collected monthly and during URI and AOM episodes from 139 infants were analyzed.16SrRNAV4 gene regions were sequenced to determine the interactions between nasopharyngeal microbial pathogens and commensals during viral upper respiratory tract infection (URI) and acute otitis media (AOM).URI frequencies were positively associated with increasing trend in otopathogen colonization.AOM frequencies were associated with decreasing trend in Micrococcus colonization.During URI and AOM, there were increases in abundance of otopathogen genera and decreases in Pseudomonas, Myroides, Yersinia, and Sphingomonas.Otopathogen abundance was increased during symptomatic viral infection, but not during asymptomatic infection.The risk for AOM complicating URI was reduced by increased abundance of Staphylococcus and Sphingobium.(67)Inconcordance with this rationale, the results of our systematic review with NMA confirms what other publication had concluded about the safety and efficacy of different probiotics as therapeutic and/or preventive options for URIs in children, adults and elderly population.