Synbiotics, probiotics or prebiotics in infant formula for full term infants: a systematic review

Background Synbiotics, probiotics or prebiotics are being added to infant formula to promote growth and development in infants. Previous reviews (2007 to 2011) on term infants given probiotics or prebiotics focused on prevention of allergic disease and food hypersensitivity. This review focused on growth and clinical outcomes in term infants fed only infant formula containing synbiotics, probiotics or prebiotics. Methods Cochrane methodology was followed using randomized controlled trials (RCTs) which compared term infant formula containing probiotics, prebiotics or synbiotics to conventional infant formula with / without placebo among healthy full term infants. The mean difference (MD) and corresponding 95% confidence intervals (CI) were reported for continuous outcomes, risk ratio (RR) and corresponding 95% CI for dichotomous outcomes. Where appropriate, meta-analysis was performed; heterogeneity was explored using subgroup and sensitivity analyses. If studies were too diverse a narrative synthesis was provided. Results Three synbiotic studies (N = 475), 10 probiotics studies (N = 933) and 12 prebiotics studies (N = 1563) were included. Synbiotics failed to significantly increase growth in boys and girls. Use of synbiotics increased stool frequency, had no impact on stool consistency, colic, spitting up / regurgitation, crying, restlessness or vomiting. Probiotics in formula also failed to have any significant effect on growth, stool frequency or consistency. Probiotics did not lower the incidence of diarrhoea, colic, spitting up / regurgitation, crying, restlessness or vomiting. Prebiotics in formula did increase weight gain but had no impact on length or head circumference gain. Prebiotics increased stool frequency but had no impact on stool consistency, the incidence of colic, spitting up / regurgitation, crying, restlessness or vomiting. There was no impact of prebiotics on the volume of formula tolerated, infections and gastrointestinal microflora. The quality of evidence was compromised by imprecision, inconsistency of results, use of different study preparations and publication bias. Authors’ conclusions There is not enough evidence to state that supplementation of term infant formula with synbiotics, probiotics or prebiotics does result in improved growth or clinical outcomes in term infants. There is no data available to establish if synbiotics are superior to probiotics or prebiotics.


Background
The first year of life is characterized by very rapid growth. Weight increases by 115%, body length 34% and head circumference 22% [1,2]. Many full term infants lose weight after birth and take 8-10 days to regain it back. The average infant achieves a weight gain of approximately 1.1 to 1.2 kg/month during the first 6 months, slowing down to 0.4 to 0.5 kg/month during the second 6 months. Length increases by 3.5 to 3.9 cm/ month during the first 4 months, slowing down to 1.8 cm/month at 6 month of age [1]. At birth average head circumference is 35 cm and increases by an estimated 12 cm during the first year of life to approximately 47 cm. A faltering head circumference has serious implications for neural growth, maturation and is diagnostic for possible problems of brain growth [2]. Monitoring growth (weight, length and head circumference) evaluates the overall health of the infant and determines adequacy of nutritional intake [1].
To promote optimum growth, development and decrease infections, probiotics, prebiotics are added to infant formula to promote an intestinal micro flora resembling that of breastfed infants [3]. The intestinal micro flora of breastfed infants have a greater concentration of bifidobacteria and fewer potentially pathogenic bacteria compared to formula fed infants. Probiotics are "live microorganisms" which when administered in adequate amounts confer a health benefit to the host [3]. The main probiotic organisms used worldwide belong to the genera Lactobacillus and Bifidobacteria and are found in the gastrointestinal micro flora [3,4]. Probiotics are consumed in the form of fermented food, dairy products, infant and toddler formula. Prebiotics are nondigestible food ingredients that benefit the host by selectively stimulating the growth and/or activity of one or a limited number of bacteria in the colon and thereby improving the host's health [4,5]. The most widely studied prebiotics are inulin, fructooligosaccharide (FOS) and galactooligosaccharide (GOS) which are plant storage carbohydrates in vegetables, cereals and fruit. Fructooligosaccharide and inulin are added to different foods as fat and sugar replacements to improve texture or for their functional benefits [5][6][7][8].
Probiotics improve health in different ways [3,9]. The health benefits conferred by probiotic bacteria are strain specific [3,9]. Some strains increase phagocytic activity of peripheral blood leukocytes, others strains promote production of mucosal antibodies reducing the transmucosal transfer of antigens. This strengthens the mucosal barrier function [10][11][12]. Other probiotic strains increase cytokine production such as interleukin 6 (IL-6) [13]. In healthy people probiotics rarely cause disease. The risk of developing bacteraemia from ingested lactobacilli is less than 1 per 1 million users; risk of developing fungaemia (from Saccharomyces Boulardii) is less than1 per 5.6 million users [14][15][16]. In many studies on infants, C-reactive protein (CRP) and IL-6 have been used to diagnose the early onset of infection [17,18]. CRP is an acute phase protein, blood levels begin to rise to 10 -1000 fold from 1 ug/ml within 4-6 hours at the onset of an infective or inflammatory process. C-reactive protein has a relatively short half-life making it useful in monitoring infection, inflammation and response to treatment [19]. IL-6 is a pro-inflammatory cytokine which stimulates the production of acute phase proteins (such as CRP) [20]. It is readily detected in serum during inflammation and indicates the presence of infection [18,19].
Adding prebiotics to formula stimulates the growth of beneficial bacteria (such as bifidobacteria, lactobacilli) in the gastrointestinal tract to levels found in breastfed infants [9,21]. As these beneficial bacteria increase, they occupy more of the "microbiological niches" in the intestine excluding pathogens. This improves the gut mucosal barrier, prevents infections with enteric pathogens or trans-located gut bacteria [22,23]. Prebiotics have a good safety record at levels found in existing food components. Flatulence or abdominal bloating is reported at doses greater than 20g / day. Abdominal cramps or diarrhoea are reported at doses greater than 50 g / day [23].
When probiotics and prebiotics are administered simultaneously, the combination is termed Synbiotics. The prebiotic in the synbiotic mixture improves the survival of the probiotic bacteria and stimulates the activity of the host's endogenous bacteria [9,21,24,25]. The superiority of synbiotics compared to either probiotics or prebiotics have not been demonstrated. No review has examined the impact of synbiotics on clinical outcomes in formula fed term infants. Recent systematic reviews (published from 2007 to 2011) on the use of probiotics or prebiotics in term infants have focused on prevention of allergic disease and food hypersensitivity [26,27]. Reviews on children and adults focused on upper respiratory tract infections, antibiotic associated diarrhoea and acute infectious diarrhoea [28][29][30]. This review focused on full term infants given only infant formula with synbiotics, probiotics or prebiotics.
The Human Research Ethics Committee at the University of Stellenbosch, South Africa reviewed the protocol, ruled that all data to be collected for this review was from the public domain and was therefore exempt from ethical approval.

Objectives
The objectives of this systematic review were: 1) To determine the effects of infant formula containing synbiotics, probiotics or prebiotics on clinical outcomes in full term infants. 2) To explore if synbiotics are superior over probiotics or prebiotics.

Criteria for considering studies for this review
All randomized controlled trials (RCTs), irrespective of language, which compared the use of term infant formula containing synbiotics, probiotics or prebiotics to conventional infant formula with or without placebo amongst healthy full term infants (>37 weeks gestation or ≥ 2.5 kg birth weight, age: 0-12 months, with no disease, congenital abnormality, allergy or eczema) receiving formula feeds only. Studies published as abstracts were included if sufficient information could be obtained to assess study quality and obtain relevant study findings.

Search methods for identification of studies
(synbiotic* and probiotic* OR prebiotic*) AND (FOS or fructooligosaccharide or inulin or GOS or galactooligosaccharide) AND (infant formula* OR infant feeding OR formula OR formula milk) AND (infant* or baby or babies) NOT (preterm or premature or low birth weight babies or allergy or eczema) AND (randomized controlled trial* OR controlled clinical trial* Or random allocation*) Limits: Humans.
We also conducted a hand search on abstracts of major conference proceedings such as the Pediatric Academic Society meetings from 1990 (www.pas-meetings. org, www.abstracts2view.com), cross checked references cited in RCTs and in recent reviews (published from 2005 to 2009) for additional studies not identified by electronic searches and specialty journals which were not included in any database such as Pediatrika and Chinese Journal of Microecology.
To identify on-going and unpublished trials, we contacted experts in the field, manufacturers of infant formula containing probiotics and prebiotics, we searched web sites of companies that have conducted or were conducting RCTs on probiotics and prebiotics e.g. Pfizer (www.pfizerpro. com/clinicaltrials), Chris Hansen Laboratory (www.chrhansen.com/research_development/documentation.html). We also searched prospective trial registries such as World Health Organization (WHO) International Clinical Trials Registry Platform Search Portal (www.who.int/ trialsearch), Clinical Trials.gov register (www.clinicaltrials.gov), Current Controlled Trials metaRegister of Controlled Trials [mRCT] (www.controlled-trials.com/ mrct) and www.clinicaltrialresults.org.

Selection of studies
One reviewer (MM) independently reviewed all abstracts, citations and identified potentially eligible studies. The full reports of eligible studies were retrieved by one reviewer (MM) and the pre-specified selection criteria applied independently by two reviewers (MM, ML) using a study eligibility form ( Figure 1). If more than one publication of a study existed, all reports of the study were grouped together under one study name. Any disagreements between the reviewers were resolved through discussion. Unresolved disagreements were resolved by a third party. Trial authors were contacted if eligibility was unclear.

Assessment of quality of evidence
Two reviewers (MM, ML) independently assessed the risk of bias of included studies as described in the Cochrane Handbook for Systematic Reviews for Interventions according to the following 6 components: 1) allocation sequence generation; 2) allocation concealment; 3) blinding; 4) incomplete outcome data; 5) selective outcome reporting; and 6) other sources of bias [31]. Where necessary, trial authors were contacted for clarification on the methodology of their studies. Any disagreements regarding risk of bias were resolved through discussion between MM, ML and RB. The quality of evidence was assessed using guidelines from the Grading of Recommendations Assessment, Development and Evaluation Working Group (GRADE), www.gradeworkinggroup. org (accessed 2012-06-07).

Data extraction and management
Two reviewers (MM, ML) independently extracted data using a pretested data extraction form. The reviewers (MM, ML) cross checked data and resolved any differences through discussion. One reviewer (MM) entered the data in Review Manager (RevMan 5) and the other reviewers (AM, ML) validated the data. Trial authors were contacted for missing data or for clarification.  Figure 1 Study Eligibility form.

Data synthesis and management
Results for probiotic, prebiotic and synbiotics studies were analysed separately. No imputation measures for missing data were applied. Trial authors were contacted if there was missing data. Available case analysis was used where there was missing data. The potential impact of missing data on results is addressed in the discussion section. Heterogeneity of the trials used in the review was assessed by visually inspecting the forest plots to detect overlapping confidence intervals and by performing a Chi 2 test (p<0.1 was considered statistically significant because of the low statistical power of this test). An I-square test (I 2 ) was also used to test for inconsistencies across studies. If the I 2 exceeded 50% and visual inspection of the forest plot supported these results, this represented substantial heterogeneity. Since all of our meta-analyses had less than ten studies, the assessment of publication bias using funnel plots could not be done [31]. If the included studies were not clinically diverse and had similar outcome measures, a meta-analysis was carried out in Review Manager (RevMan 5) by two reviewers (AM, MM). The random effects meta-analysis model was applied to all meta-analyses since the studies were clinically heterogeneous in terms of different settings (countries), doses and strains of synbiotics, probiotics or type of prebiotics, different treatment durations, and other unforeseen factors. The inverse-variance method was used for continuous data and the Mantel-Haenszel method was used for dichotomous data. For continuous outcomes the mean difference (MD) and corresponding 95% confidence intervals (CI) were calculated. For dichotomous outcomes, the risk ratio (RR) and corresponding 95% CI were calculated. The source of statistical heterogeneity was explored using subgroup and sensitivity analyses. If studies were too diverse, no meta-analysis was conducted and a narrative synthesis was provided.

Results of the search and description of studies
Electronic search of available databases yielded 142 citations. After reading titles and abstracts, duplicate reports were removed, 118 articles were screened and 55 articles were excluded. A hand search yielded 2 more articles. Potentially relevant full text reports were retrieved, reviewed for eligibility and a further 38 studies were excluded. One study was published in two other reports [32][33][34]. The three studies were considered as one study and are referred to as Moro 2006 [32]. Another study was also published in two reports; and is referred as Moro 2002 [35,36]. Twenty five studies (3 synbiotic, 10 probiotic and 12 prebiotic studies) and three on-going studies were included in this review [21,24,25,[37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56]. The selection process is shown in Figure 2. Table 1 gives a list of 38 studies which were excluded for: use of breast milk or mixed feeds (12 studies), no use of probiotic or prebiotic (2 studies), being a cross over study, not RCT (5 studies), type of feed was unspecified (3 studies), different inclusion criteria or outcomes (12 studies), no data available for end of treatment period (1 study) and data presentation inappropriate for meta-analysis (3 studies) . No eligible studies were excluded for failure to report the review's pre-specified outcomes.

Included studies
Summary of the included synbiotics, probiotics, prebiotics, and on-going studies are shown in Tables 2 3, 4, 5. All studies were conducted on healthy infants and used standard infant formula.
Probiotic studies: Ten probiotic studies (N = 933) were included. One study [55] used a reduced protein infant formula and one study [50] used an acidified formula given to healthy infants born to HIV positive mothers. The most widely studied probiotics were Bifidobacterium lactis (BB-12) which was administered alone [40,44,46,[50][51][52]. Other probiotic strains used were Lactobacillus reuteri and Bifidobacterium bifidum. Doses ranged widely. For Bifidobacteria: 1.5 x 10 6 to 3.85 x 10 8 cfu/g powder and Lactobacillus: 1 x 10 6 to 1 x 10 8 cfu/g powder. Treatment duration varied from 14 days to 7 months. The probiotic studies were conducted in Australia (Adelaide), Belgium, Chile (Santiago), France, Israel (Beersheva), South Africa (Johannesburg) and USA (Iowa). None of the probiotic studies reported data on immune response.

Primary outcomes
Growth parameters (i) Weight gain Only one study [24] reported weight gain in terms of grams per day (g/day). In this study, two types of synbiotics (Type 1 and Type 2) were evaluated and results for boys and girls were reported separately. The results of the two synbiotics were combined using the One study [21] reported weight gain in terms of some score scale. A calculated treatment effect showed that synbiotics failed to significantly increase weight gain compared to controls (MD −0.07, 95% CI: -0.43 to 0.29, n = 79). Since the score scale can take negative values, the values of mean and standard deviation in this analysis do not necessarily imply that the data is skewed.
One study [25] reported weight gain (g/day) in terms of mean difference (MD) and 90% CI. These values were used in calculating the corresponding standard error (SE  were combined before meta-analysis. Results from the meta-analysis showed that synbiotics failed to significantly increase head circumference gain compared to controls for both boys. There was no significant heterogeneity detected between the two studies for both boys (Chi 2 =0.64, df=1, p=0.43, I 2 =0%) and girls (Chi 2 =0.67, df=1, p=0.41, I 2 =0%).
One study [21] reported head circumference gain in terms of some score scale. A calculated treatment effect showed that synbiotics failed to significantly increase head circumference gain compared to controls (MD 0.01, 95% CI: -0.38 to 0.36, n = 79). Since the score scale can take negative values, the values of mean and standard deviation in this analysis do not necessarily imply that the data is skewed.
One study [25] reported stool frequency (evacuations per day) but values for standard deviations were not given and as a result, no treatment effect could be calculated. (ii) Stool consistency One study [21] evaluated stool consistency using a consistency score (1=hard to 4=watery and loose) and a calculated treatment effect showed no significant difference between the synbiotic and control treated groups (MD 0.13, 95% CI: -0.15 to 0.41, n = 79).
One study [24] study reported that liquid stools occurred significantly more frequently in the synbiotic group compared to the control group (OR 3.17, 95% CI: 1.59 to 3.60, n = 66).
Puccio 2007 [25] reported that data on stool consistency showed no statistically significant differences between the two study groups (data not shown in study report). (iii) Incidence of colic, spitting up / regurgitation, vomiting, crying Data on frequency of crying, restlessness, colic, spitting and vomiting reported by Puccio 2007 [25] showed no statistically significant differences between the two study groups (data not shown in study report). Results from Vlieger 2009 [21] showed no significant differences in the frequency of vomiting (RR 0.46, 95% CI: 0.12 to 1.72, n = 79) and colic (RR 2.50, 95% CI: 0.46 to 13.73, n = 79) between the two study groups. The same study showed no difference in crying (hours per day) between the two study groups (MD −0.10, 95% CI: -0.46 to 0.26, n = 79). (iv) Average formula intake One study [25] reported the mean daily intake of formula in a graph where no values could be retrieved.  One study [25] reported the mean daily intake of formula in a graph where no values could be retrieved.

(i) Infections
Puccio 2007 [25] reported data on frequency of respiratory tract infections but there were no significant differences between the synbiotic and control treated groups (RR 0.71, 95% CI: 0.31 to 1.59, n = 97). Vlieger 2009 [21] reported the mean (SD) of upper respiratory tract infections and gastrointestinal infections (times per month) but no treatment effect could be calculated because the data was skewed (mean < SD). (ii) Antibiotic intake Vlieger 2009 [21] reported the mean (SD) of the use of antibiotics (times per month) but no treatment effect could be calculated because the data was skewed (mean < SD).

(ii) Length gain
Four studies [24,44,50,55] reported length gain for boys and girls separately. Two studies reported in terms of mm/month and two studies reported in terms of mm/day. The latter two studies results were converted to mm/month by multiplying both the mean and SD by 28, assuming a 4 week/ 28-day month. Results from these four   [55] reported that that there was no significant formula effects on crying and colic (data not shown in report). (v) Average formula intake One study [38] reported the average formula intake (ml/kg body-weight /day) and the calculated treatment effect showed no differences between the probiotic and control groups (MD 5.00, 95% CI: -12.60 to 22.60, n = 58). Two studies [44,51] reported the average formula intake (ml/day) and meta-analysis showed that infants in the probiotic group had a significantly higher formula intake compared to the controls (MD 46.74, 95% CI: 23.93 to 69.54, n = 292). No statistically significant heterogeneity was detected between the studies (Chi 2 =0.45, df=1, p=0.50, I 2 =0%).

(i) Infections
One study [44] reported the number of infants having respiratory infections and the calculated treatment effect showed no differences between the probiotic and control groups (RR 0.93, 95% CI: 0.74 to 1.17, n = 142). One study [51] reported episodes of respiratory illness in terms of mean (95% CI). The mean (95% CI) were used in calculating the SDs. However, no treatment effect was calculated because the data was skewed (mean < SD). One study [44] reported the number of infants having gastrointestinal infections and the calculated treatment effect showed no differences between the probiotic and control groups (RR 0.70, 95% CI: 0.45 to 1.11, n = 142). (ii) Antibiotic intake One study [51] reported prescription of antibiotics in terms of mean (95% CI). The mean (95% CI) were used in calculating the SDs. However, no treatment effect was calculated because the data was skewed (mean < SD).

Hospitalization
Only one study [55] reported hospitalization but no treatment effect was calculated because the data was skewed (mean < SD)

Prebiotics versus controls
Twelve studies (N = 1563) investigated the effect of prebiotic administration versus placebo or no prebiotic in formula (Control group) [ The results for the prebiotics in each of these studies were combined before meta-analysis using combined mean and pooled standard deviation ( Figure 8).

(ii) Length gain
Seven studies [32,35,41,42,47,49,54] reported length gain either as cm/week or in units that were converted to cm/week. Meta-analysis of their results showed that prebiotics failed to significantly increase length gain compared to the controls (MD 0.01, 95% CI: -0.01 to 0.04, n = 697). No statistically significant heterogeneity was detected     [35] were combined before meta-analysis using combined mean and pooled standard deviation.

(i) Infections
According to Moro 2006 [32], prebiotics significantly reduced overall infections compared to the controls (RR 0.45, 95% CI: 0.29 to 0.69, n = 204). The number of infants having gastrointestinal infections, urinary tract infections (UTI) and otitis media were very few [32]. Two studies [32,39] reported the number of infants with upper respiratory tract infections (URTI) and their results were pooled in a meta-analysis. However, due to significant heterogeneity detected between the two studies (Chi 2 =7.69, df=1, p=0.006, I 2 =87%), their results are reported separately. Although Moro 2006 [32] showed that the prebiotic group significantly reduced the number of infants with URTI compared to the controls (RR 0. 48 [42] were combined before meta-analysis using combined mean and pooled SD. However, Brunser 2006 [38] showed no significant difference in the number of bifidobacteria between the two groups (MD −0.39, 95% CI: -1.49 to 0.71, n = 43) [ Figure 9]. Four studies reported their results in median; therefore no conclusions could be made. Costalos 2008 [41] reported the median (range) of Bifidobacteria (log10 CFU per gram stool) as a percentage of total bacteria (Table 7). Three studies [32,35,43] reported the median (IQR) of Bifidobacteria (log10 CFU per gram stool) ( Table 8).

(ii) Lactobacillus
Three studies [38,53,54] reported Lactobacillus (log10 CFU per gram stool) and meta-analysis of their results showed statistically significant heterogeneity between the studies (Chi 2 =26.44, df=2, p < 0.00001, I 2 =92%). Sensitivity analysis was done by removing the one study [38]  Two studies [53,54] reported E. coli (log10 CFU per gram stool) and their results were pooled in a metaanalysis. However, statistically significant heterogeneity was detected between the studies (Chi 2 =5.96, df=1, p=0.01, I 2 =83%). The results are therefore reported separately. Xiao-Ming 2004 [53] showed that prebiotics significantly reduced E. coli counts compared to the controls (MD −0.90, 95% CI: -1.29 to −0.51, n = 121) while Xiao-Ming 2008 [54] showed no significant difference between the two groups (MD 0.67, 95% CI: -0.53 to 1.87, n = 38).   *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the measure of effect of the intervention (and its 95% CI).CI: Confidence interval, MD: Mean Difference. GRADE Working Group grades of evidence. High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. 1 Small sample size n=81, 95% CI includes no effect. 2 Possible publication bias. 3 Small sample size n=86, 95% CI includes no effect. 4 Possible publication bias. 5 Allocation concealment not described in 2 studies. 6 Small sample size n=126. 7 Possible Publication bias. 8 Allocation concealment not described in 2 studies. 9 Small sample size n=138. 10 Possible Publication bias. 11 Small sample size n=126. 12 Possible publication bias. 13 Small sample size n=138. 14 Possible publication bias. 15 Small sample size n=176. 16 Possible publication bias.
Two studies reported their results in median; therefore no conclusions could be made. Costalos 2008 [41] reported the median (range) of E. coli (log10 CFU per gram stool) as a percentage of total bacteria (Table 8). Fanaro 2008 [43] reported the median (IQR) of E. coli and clostridia (log10 CFU per gram stool) ( Table 8).

Discussion
The objectives of this systematic review were to determine the effects of infant formula containing probiotics, prebiotics or both (synbiotics) on clinical outcomes in full term infants and to explore if synbiotics are superior over probiotics or prebiotics. Studies that used breast milk or mixed feeds (breast and infant formula or other types of milk) were excluded. All included RCTs evaluated either synbiotics, probiotics or prebiotics use in full term infants. The studies varied in enrolment criteria, sample size, intervention and treatment duration.

Summary of main findings Synbiotics
Addition of synbiotics to infant formula did not have any significant effect on growth (weight gain, length and head circumference). Synbiotics significantly increased stool frequency. However, two studies [21,25] reported no differences in stool consistency, while one study [24] reported an increase in liquid stools in synbiotic group. There were no significant differences between study groups on the incidence and frequency of colic, spitting up / regurgitation, crying, restlessness or vomiting. The effect of synbiotics on the volume of formula tolerated was not reported. Effect of synbiotics on frequency of infections was under reported. In one study [25], there were no differences in the frequency of infections between study groups, while in another study [21], the treatment effect could not be calculated or any conclusions made on the frequency of infections or antibiotic intake. Effects of synbiotics on hospitalization, GI microflora and immune response were not reported in any study therefore these parameters could not be evaluated.
Interpreting the effects of synbiotic supplementation of infant formula on clinical outcomes was difficult due to the limited number of studies. The synbiotic studies had short treatment duration (4 to 6 months) and treatment varied in all 3 studies. There was not enough evidence to state that synbiotics in infant formula have a significant effect on growth or lower the incidence of colic, spitting up / regurgitation, crying, restlessness. There is limited evidence that synbiotics do increase stool frequency and effects on stool consistency were inconclusive. There is not enough evidence to state that synbiotics reduce the risk of infections or decrease use of antibiotics. There is no data on the effects of synbiotics on GI microflora. The available data is very limited to draw reliable conclusions on the effects of synbiotics on clinical outcomes in formula fed infants.   *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the measure of effect of the intervention (and its 95% CI). CI: Confidence interval, CFU: colony forming units, MD: Mean Difference, RR: Risk ratio. GRADE Working Group grades of evidence. High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. 1 Small sample size n=158, 95% CI includes no effect. 2 Possible publication bias. 3 Unexplained heterogeneity). 4 Small sample size n=170. 5 Possible publication bias. 6 Small sample size n=158, 95% CI includes no effect. 7 Possible publication bias. 8 Small sample size n=165, 95% CI includes no effect. 9 Possible publication bias. 10 Small sample size n=125, 95% CI includes no effect. 11 Possible publication bias. 12 Small sample size n=139. 13 Possible publication bias. 14 Small sample size n=57. 15 Possible publication bias.

(3 studies) ÈÈ⊝⊝
low 19,20,21 Follow-up: 3 to 6 months *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the measure of effect of the intervention (and its 95% CI). CI: Confidence interval, CFU: Colony Forming Units, MD: Mean Difference, RR: Risk ratio. GRADE Working Group grades of evidence: High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. 1 Allocation concealment not clearly described in 6 studies. 2 Blinding not clearly demonstrated or described in 7 studies. 3 Possible publication bias. 4 Allocation concealment not clearly demonstrated in 5 studies. 5 Blinding not clearly demonstrated in 6 studies. 6 Possible publication bias. 7 Blinding not clearly described in 2 studies. 8 Possible publication bias. 9 Incomplete outcome data (with no reasons given for missing data) was present in 1 study. 10 Possible publication bias. 11 Small sample size n=237, 95% CI includes no effect. 12 Possible publication bias. 13 Unexplained heterogeneity. 14 95% CI includes no effect. 15 Possible publication bias. 16 Unexplained heterogeneity. 17 Small sample size n=280. 18 Possible publication bias. 19 Unexplained heterogeneity. 20 Small sample size n=202. 21 Possible publication bias.

Probiotics
A limited number of studies analyzed the effects of probiotic supplementation on growth by gender. These studies had small sample sizes and the follow-up periods were short. Addition of probiotics to infant formula did not have any significant effect on growth (weight gain, length gain or head circumference) in boys or girls. No study reported any weight loss. Probiotic infant formula was well tolerated. The limited available data shows that probiotics did not have any significant effect on stool frequency or consistency. Probiotic supplementation was not associated with fewer episodes of diarrhoea, a lower incidence of colic, spitting up / regurgitation, restlessness, vomiting. In one study [52] there were fewer crying episodes in the control group than probiotic group. Probiotic effects on infections, antibiotic use and length of hospitalization were inconclusive. Probiotic supplementation did result in a significantly higher formula intake compared to controls.
Effects of probiotic supplementation on intestinal microflora were conflicting. Probiotics failed to increase counts of bifidobacteria and lactobacillus. Probiotics significantly reduced counts of enterobacteria but failed to reduce counts of bacteriodes. None of the studies reported on immune response (CRP, IL-6), therefore the impact of probiotics on these parameters could not be evaluated. All 10 probiotic studies used various strains of bifidobacteria and lactobacillus with different doses. Treatment duration also varied from 14 days to 7 months. This confirms the ESPGHAN Committee on nutrition statement that there is a lack of published evidence on clinical benefits from long term use of probiotic containing infant formula [95]. Well designed long term followup RCTs using similar treatment regimens (same probiotics strains, dose and treatment duration) are needed to establish the effects of probiotics on healthy formula fed infants.

Prebiotics
Prebiotic addition to infant formula did have a significant effect on weight gain but had no significant effect on length and head circumference. None of the studies reported any weight loss. Prebiotic supplementation increased stool frequency but failed to improve stool consistency or decrease incidence of diarrhoea. Prebiotic supplementation did not reduce the incidence of spitting up / regurgitation, vomiting or crying (no study reported colic) or increased volume of formula tolerated. Prebiotic supplementation failed to significantly reduce upper respiratory infections. However, one study [32] did report a significant reduction in overall infections and antibiotic intake. Prebiotics supplementation failed to increase counts of bifidobacteria, lactobacillus or decrease the levels of pathogens (enterobacteria, bacteriodes, Ecoli). None of the studies reported on hospitalization (days in hospital) and immune response (CRP, IL-6), therefore the impact of prebiotics on these parameters could not be evaluated.
Majority of the studies had a short treatment duration ranging from 28 days to 12 months. The prebiotic doses ranged from 0.15 g to 0.8 g/100 ml which did not exceed the level recommended by the European Committee on food in order to minimize intolerance and maximize the bifidogenic effect of the prebiotic.

Quality of the evidence and potential biases in the review process
We used guidelines from GRADE working group and GRADEpro 3.6 software to assess the quality of evidence in this review ( Table 7, 9,10). Overall the quality of evidence for primary outcomes is low, meaning that further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. The quality of the evidence was compromised by: Imprecision (majority of studies had a small sample size ranging from 97 to 227 in the synbiotic studies, 54 to 201 in probiotic studies, 32 to 271 in the prebiotic studies); limitations in study design and execution (inadequate information was published to assess methodological quality of the study. Information was missing on sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, free of other bias domains; Inconsistency of results); unexplained heterogeneity; use of different study preparations (types of synbiotic, probiotic, prebiotics) and different doses regimens were used and publication bias.
At the conclusion of the review process and preparation of the manuscript (for this review), one on-going study [96] was recruiting, one study [97] was not yet recruiting, one study [98] was still on-going, no longer recruiting. Therefore data from these studies could not be included in this review. The reviewers used thorough comprehensive search strategies adopted for the available databases. All attempts were made to minimize publication bias. All steps of this review were conducted independently by the reviewers. Only randomised controlled studies were included in this review.

Breast feeding statement
By conducting this review on exclusively formula fed infants, the authors do not seek to diminish the importance of breastfeeding and promote formula feeding. The reviewers acknowledge the importance of breastfeeding for infants. They support exclusive breastfeeding for 6 months, thereafter safe complementary feeding from 6 months of age with continued breastfeeding up to 2 years and beyond as per the global recommendations for optimal infant feeding of WHO and United Nations Children's Fund (UNICEF). This is because breastfeeding is the ideal feeding method for infants [99].

Conclusion
There is not enough evidence to state that supplementation of term infant formula with synbiotics, probiotics or prebiotics does result in improved growth and clinical outcomes in full term infants. There is no data available to establish if synbiotics are superior to probiotics or prebiotics. Therefore this review does not support the routine supplementation of term infant formula with synbiotics, probiotics or prebiotics.

Implications for practice
Probiotics: The limited evidence shows synbiotic or probiotic supplementation of infant formula did not have any adverse effects, significant impact on growth or clinical outcomes. All studies used different probiotic strains, the effects of one type of probiotic cannot be extrapolated to other types of probiotic bacteria. Prebiotic supplementation of infant formula also did not result in any adverse effects on infants. There are some clinical benefits such as improved weight gain and stool frequency.

Implications for research
For clear recommendations to be made, well designed large RCTs with long term follow -up are required on exclusively formula fed term infants to investigate the effect of the same synbiotic combinations on clinical outcomes; the effect of the same probiotics (with similar doses and treatment duration) on clinical outcomes because available studies used different probiotic doses and treatment durations; the effect of the same prebiotics (with similar doses and treatment duration) on clinical outcomes because available studies used similar prebiotics with different doses and treatment duration; the effects of synbiotics, probiotics or prebiotics on clinical outcomes that have not been adequately addressed in previous studies; if synbiotics are superior to probiotics or prebiotics. Future RCTs should have treatment arms that include both synbiotics, probiotic and prebiotics.