Probiotics and Time to Achieve Full Enteral Feeding in Human Milk-Fed and Formula-Fed Preterm Infants: Systematic Review and Meta-Analysis

Probiotics have been linked to a reduction in the incidence of necrotizing enterocolitis and late-onset sepsis in preterm infants. Recently, probiotics have also proved to reduce time to achieve full enteral feeding (FEF). However, the relationship between FEF achievement and type of feeding in infants treated with probiotics has not been explored yet. The aim of this systematic review and meta-analysis was to evaluate the effect of probiotics in reducing time to achieve FEF in preterm infants, according to type of feeding (exclusive human milk (HM) vs. formula). Randomized-controlled trials involving preterm infants receiving probiotics, and reporting on time to reach FEF were included in the systematic review. Trials reporting on outcome according to type of feeding (exclusive HM vs. formula) were included in the meta-analysis. Fixed-effect or random-effects models were used as appropriate. Results were expressed as mean difference (MD) with 95% confidence interval (CI). Twenty-five studies were included in the systematic review. In the five studies recruiting exclusively HM-fed preterm infants, those treated with probiotics reached FEF approximately 3 days before controls (MD −3.15 days (95% CI −5.25/−1.05), p = 0.003). None of the two studies reporting on exclusively formula-fed infants showed any difference between infants receiving probiotics and controls in terms of FEF achievement. The limited number of included studies did not allow testing for other subgroup differences between HM and formula-fed infants. However, if confirmed in further studies, the 3-days reduction in time to achieve FEF in exclusively HM-fed preterm infants might have significant implications for their clinical management.


Introduction
Nutrition during critical time windows in early life can affect long-term health [1]. Early provision of optimal enteral nutrition to preterm infants might improve neurodevelopmental outcome by decreasing the rate of several complications of prematurity, such as extrauterine growth restriction, necrotizing enterocolitis (NEC), sepsis, bronchopulmonary dysplasia, and retinopathy of prematurity [2].
Late introduction and slow advancement of enteral feeding may alter gastrointestinal motility and disrupt microbial colonization [3], leading to a delay in establishing full enteral feeding (FEF). The consequent prolonged need for parenteral nutrition can have serious infectious and metabolic complications, which might prolong hospital stay, increase morbidity and mortality, and affect growth and development [4].
Several clinical variables and interventions have been proposed as predictors of the time to FEF achievement in preterm and very-low-birth-weight (VLBW) infants. Among these variables, the influence of type of feeding was also documented, as FEF achievement was delayed in formula-fed infants compared to human milk (HM)-fed infants [5].
Recently, probiotic use has been associated with a reduced time to achieve FEF and better feeding tolerance [6], as well as a reduction of NEC [7,8] and late-onset sepsis [9]. Probiotics are live microorganisms which, when ingested in adequate amounts, confer a health benefit to the host, by modifying the composition and function of gut microbiota and the immunological responses in the host [10]. The role of probiotics in attaining a more rapid achievement of FEF could be related to their favorable effect on the physiological intestinal dysbiosis of preterm infants [11], which is the result of the exposure to a unique environment and to several iatrogenic manipulations, such as broad spectrum antibiotics [12]. It is well known that gut microbiota in HM-fed infants is different compared to formula-fed infants [13]; data from an observational study also suggest a feeding-dependent effect of probiotics, as in that study NEC incidence was reduced in infants treated with probiotics and receiving HM, but not in those exclusively formula-fed [14]. However, the relationship between probiotics and type of feeding in attaining a more rapid achievement of FEF has not been explored yet, even in the most recent meta-analysis on this topic [6].
Thus, the aim of the present paper was to evaluate the effect of probiotics on time to FEF achievement according to type of feeding (exclusive HM vs. formula), by performing a systematic review and meta-analysis of currently available literature on this topic.

Literature Search
The study protocol was designed by the members of the Task Force on Probiotics of the Italian Society of Neonatology. PRISMA guidelines [15] were followed in order to perform a systematic review of published studies reporting the relationship between probiotic use and time to FEF achievement in preterm infants according to type of feeding.
In order to be included in the meta-analysis, studies had to meet the following inclusion criteria: randomized or quasi-randomized clinical trials involving preterm infants (gestational age (GA) <37 weeks) who received, within one month of age, any probiotic compared to placebo or no treatment, and reporting on type of feeding. The outcome of interest was time for FEF achievement (any definition). Only English-written studies and studies involving humans were included in the meta-analysis.
A search was conducted for studies published before 2 March 2016 in PubMed [16], the Cochrane Library [17], and Embase [18]. The following search string was used for the PubMed search: ((preterm infant OR pre-term infant) OR (preterm infants OR pre-term infants) OR (preterm neonate OR pre-term neonate) OR (preterm neonates OR pre-term neonates) OR (preterm newborn OR pre-term newborn) OR (preterm newborns OR pre-term newborns) OR (premature infant OR premature infants) OR (premature neonate OR premature neonates) OR (premature newborn OR premature newborns) OR infant, extremely premature (MeSH Heading (MH)) OR premature birth (MH) OR infant, low birth weight (MH) OR infant, very low birth weight (MH)) AND (full enteral* OR feed*) AND (probiotic OR probiotics OR pro-biotic OR pro-biotics OR probio*)) NOT (animals (MH) NOT humans (MH).
The string was built up by combining all the terms related to probiotics and FEF achievement: PubMed MeSH terms, free-text words, and their combinations obtained through the most proper Boolean operators were used. The same criteria were used for searching the Cochrane Library and Embase.
Arianna Aceti and Luigi Corvaglia performed the literature search: relevant studies were identified from the abstract; full-texts of relevant studies were examined, as well as their reference lists in order to identify additional studies.

Data Extraction and Meta-Analysis
Study details (population, characteristics of probiotic and placebo, type of feeding, and outcome assessment) were evaluated independently by Arianna Aceti and Luigi Corvaglia, and checked by Davide Gori. Study quality was evaluated independently by Arianna Aceti and Davide Gori using the risk of bias tool as proposed by the Cochrane collaboration (Chapter 8 of the Cochrane Handbook of Systematic Reviews) [19].
The corresponding authors of the studies in which days to FEF achievement were not reported as mean˘standard deviation (SD) were contacted by email. When data were not provided, the study was not included in the meta-analysis.
The association between probiotic use and FEF achievement according to type of feeding was evaluated by a meta-analysis conducted by AA and DG using the RevMan software (Cochrane Informatics and Knowledge Management Department, version 5.3.5) downloaded from the Cochrane website [20]. Mean difference (MD) in days to achieve FEF between infants receiving probiotics and those receiving placebo or no treatment was calculated using the inverse variance method, and reported with 95% confidence interval (CI).
For the analysis, we planned to use at first a fixed effect model. Heterogeneity was measured using the I 2 test: if significant heterogeneity was present (p < 0.05 from the χ 2 test) and/or the number of studies was ď5, a random-effects model was used instead.
Twenty-four studies were excluded after examining the full-texts [ Among them, only eight studies reported FEF achievement according to type of feeding: infants were fed exclusively HM, either own mother's (OMM) or donor human milk (DHM), in six studies [22,38,50,56,60,70], while two studies reported FEF in exclusively formula-fed infants [41,61].
The corresponding authors of four of these papers were contacted by email, as data for FEF achievement were not suitable for inclusion in the meta-analysis: mean˘SD of days for FEF achievement were provided for one study [22], while data were unavailable for three studies [41,61,70]; these three studies were thus excluded from the meta-analysis.
Overall, five studies were included in the meta-analysis: in all these studies, infants were fed exclusively HM, either OMM or DHM ( Figure 1) [22,38,50,56,60].      Van Niekerk, 2014 [22] P Preterm infants with GA < 34 weeks and BW < 1250 g, exposed and non-exposed to HIV (only infants unexposed to HIV are included in the meta-analysis) L. rhamnosus, B. infantis  Three studies were not included in the meta-analysis because data on FEF were not available as mean ± SD [41,61,70]. One study reported the use of Bacillus clausii in preterm infants with GA < 34 weeks, fed expressed breast milk or DHM [70] and stratified as extreme preterm (GA 27-30 + 6 weeks) and very preterm (GA 31-33 + 6 weeks). In both groups, probiotic use was associated with a reduced time to achieve FEF (risk ratio 0.82 (95% CI 0.74-0.88) and 0.67 (95% CI 0.32-0.77), respectively).
The other two studies reported probiotic use in exclusively formula-fed infants: in the study by Costalos et al., infants born at 28-32 weeks gestation and fed exclusively preterm formula received Saccharomyces boulardii or placebo for approximately 30 days [41]. In the study by Stratiki et al., formula-fed infants with a similar gestational age (27-32 weeks) received Bifidobacterium lactis vs. no treatment [61]. Neither of these two studies reported any significant difference between groups in terms of time to FEF achievement.
All the studies included in the meta-analysis, except one [50], recruited exclusively infants with birth weight <1500 g. The study by Roy et al. [50] reported specific data for extremely low birth weight (ELBW) infants: time to reach FEF in ELBW infants treated with probiotics was significantly lower than in controls (mean ± SD 13.22 ± 5.04 vs. 17.41 ± 8.07, respectively, p = 0.014). None of the studies included in the meta-analysis reported separate data on intrauterine growth restricted (IUGR) infants.
In all the studies, except one [56], a probiotic mix was used: the meta-analysis performed after the exclusion of the study by Manzoni et al., where a single-strain product containing Lactobacillus GG was used, confirmed the results of the overall analysis (MD −3.33 days (95%CI −5.63/−1.04), p ≤ 0.004). Three studies were not included in the meta-analysis because data on FEF were not available as mean˘SD [41,61,70]. One study reported the use of Bacillus clausii in preterm infants with GA < 34 weeks, fed expressed breast milk or DHM [70] and stratified as extreme preterm (GA 27-30 + 6 weeks) and very preterm (GA 31-33 + 6 weeks). In both groups, probiotic use was associated with a reduced time to achieve FEF (risk ratio 0.82 (95% CI 0.74-0.88) and 0.67 (95% CI 0.32-0.77), respectively).

Methodological Study Quality
The other two studies reported probiotic use in exclusively formula-fed infants: in the study by Costalos et al., infants born at 28-32 weeks gestation and fed exclusively preterm formula received Saccharomyces boulardii or placebo for approximately 30 days [41]. In the study by Stratiki et al., formula-fed infants with a similar gestational age (27-32 weeks) received Bifidobacterium lactis vs. no treatment [61]. Neither of these two studies reported any significant difference between groups in terms of time to FEF achievement.
All the studies included in the meta-analysis, except one [50], recruited exclusively infants with birth weight <1500 g. The study by Roy et al. [50] reported specific data for extremely low birth weight (ELBW) infants: time to reach FEF in ELBW infants treated with probiotics was significantly lower than in controls (mean˘SD 13.22˘5.04 vs. 17.41˘8.07, respectively, p = 0.014). None of the studies included in the meta-analysis reported separate data on intrauterine growth restricted (IUGR) infants.
In all the studies, except one [56], a probiotic mix was used: the meta-analysis performed after the exclusion of the study by Manzoni et al., where a single-strain product containing Lactobacillus GG was used, confirmed the results of the overall analysis (MD´3.33 days (95% CI´5.63/´1.04), p ď 0.004).

Methodological Study Quality
Evaluation of the quality of the studies included in the meta-analysis according to the risk of bias tool as proposed by the Cochrane Collaboration [19] is shown in Table 2.

Discussion
The present meta-analysis shows that the use of probiotics in preterm, VLBW infants fed exclusively HM is associated with 3-days reduction in the time to FEF achievement. The only two studies included in the present systematic review in which infants were exclusively formula-fed did not report any difference between the probiotic and the control group.
The single previous meta-analysis investigating FEF as primary outcome showed an overall smaller reduction in the time to FEF achievement, but did not report separate data for HM-fed and formula-fed infants [6]. The studies included in the meta-analysis by Athalye-Jape et al. are almost the same as those included in our systematic review; quite surprisingly, in the majority of the studies included in these two reviews, both HM and formula-fed infants were recruited, but no detailed information on the relationship between type of feeding and outcome was provided.
Type of feeding might modulate the relationship between probiotics and neonatal clinical outcome [14]. It has been previously shown that HM feeding is associated with shorter time to achieve FEF compared to formula feeding [5]. Our meta-analysis, which included only studies where infants were exclusively HM-fed, showed a significant reduction in the time to achieve FEF attributable to probiotics. Despite the limitation given by the small number of studies, a probiotic-related 3-days reduction in time to achieve FEF in preterm infants fed exclusively HM has strong clinical implications and deserves further consideration. When OMM is not available or contraindicated, the use of pasteurized DHM is recommended for preterm infants: pasteurization inactivates most viral and bacterial agents, but at the same time affects some nutritional and immunological properties of HM, including endogenous probiotics [71]. It can be speculated that the beneficial effect of probiotics documented in exclusively HM-fed infants could be attributed to a synergic action exerted by the prebiotic components of HM and the exogenous probiotic, which partially restores the symbiotic properties of naïve HM [72]. In the present meta-analysis, no separate data for OMM-fed and DHM-fed infants were available; for this reason, it is not possible to clarify whether the beneficial effect of HM on FEF achievement applies both to OMM and to DHM.
Heterogeneity among included studies was high; however, given the small number of papers, our ability to explore sources of heterogeneity was limited. In the five included studies, different probiotic strains were used. We aimed to perform strain-specific sub-meta-analyses, in order to clarify whether there was any probiotic product showing a significant benefit in terms of reduction in the time to achieve FEF. However, such analyses were not feasible, as none of the studies used the same probiotic strain or mix. Similarly, it was not possible to explore additional sources of heterogeneity, such as the characteristics of probiotic administration (dose, duration, infant age at probiotic initiation, etc.). In addition, we were unable to test for subgroup differences between HM-fed and formula-fed infants, which might have partially explained the different results in terms of FEF achievement.
Apparently, studies were homogeneous in terms of included populations, as almost all of them recruited only VLBW infants. However, few data on "high-risk" infants, such as ELBW and IUGR infants, could be extrapolated from the main results of the included studies.
The use of probiotics should be weighed against their potential side effects. There are some reports about the occurrence of sepsis in preterm newborns, potentially linked to probiotic administration [73].
However, none of the studies included in the systematic review reported any side effect related to the use of probiotics.

Conclusions
According to the results of the present meta-analysis, the use of probiotics is linked to 3-days reduction in time to achieve FEF in preterm VLBW infants fed exclusively HM. If confirmed in further studies, this reduction might have strong clinical implications for this high-risk population.