Effectiveness of probiotics in infantile colic: A rapid review

Abstract

Background

Infantile colic (IC) is a troubling condition with limited treatment options for young infants. This rapid review aims to synthesize the evidence for probiotics in the treatment and prevention of IC in healthy term infants.

Methods

We searched in MEDLINE, EMBASE, CINAHL, Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews for systematic reviews (SRs), and randomized control trials (RCTs) published between January 1, 2000 and July 11, 2018. Trials were included if they recruited healthy full-term infants who received probiotics for treatment or prevention of colic. The quality of evidence was assessed using GRADE criteria. As supplementary information, the safety of probiotics in infants was searched within the reviewed studies and other recent publications.

Results

We identified four SRs and meta-analyses that included six RCTs, and found an additional three RCTs evaluating probiotics for the treatment of IC. One SR and six RCTs were identified for prevention of IC; four of the RCTs were included in the SR and two were published later. The probiotic Lactobacillus reuteri was used in the majority (five of eight) of treatment trials, and was found to significantly reduce crying in colicky breast-fed infants compared to placebo. Only two of the six prevention trials showed a significant decrease in crying time compared to placebo, although another two trials showed other benefits of probiotics, including reduced use of medications (simethicone and cimetropium bromide) and physician visits. No adverse events were identified in the included studies; other research suggests probiotics are generally safe in healthy children.

Conclusion

This rapid review identified limited but favourable evidence of benefit of using probiotics for the treatment of IC in full-term breast-fed infants. While routine use of probiotics for treating or preventing IC cannot yet be recommended, it can be an option to manage IC.

BACKGROUND

Infantile colic (IC) is a common condition in young infants, affecting up to 20% of breast- and formula-fed infants in the first 3 months of life and causing considerable distress for many families (1,2). Wessel’s criteria defines IC as paroxysms of irritability, fussing or crying lasting more 3 hours per day, occurring on more than 3 days per week, for a period of 3 or more weeks (3); modified Wessel’s criteria omits the 3-week minimum (4). IC can start as early as 2 weeks of age, usually peaks at 6 to 8 weeks, and resolves spontaneously by age 3 to 4 months (5). High pitched crying, a flushed face, apparent difficulty passing gas, abdominal distension, difficulty passing stools, drawing up the legs, and arching the back are commonly reported signs and symptoms of colic in otherwise thriving infants (5).

Despite much research, the etiology of IC remains unknown. A number of therapies have been studied, including sucrose, herbal teas, changing infant formula, swaddling, pacifiers, music, vibration or massage, and spinal manipulation. While some clinical trials have shown that elimination of cow’s milk protein from the breastfeeding mother’s diet and/or using hypoallergenic formula may decrease colic (6–8), no therapy has been proven consistently effective (9), and parents continue to seek help from physicians and other sources.

Probiotics are broadly defined as ‘live microorganisms which, when administered in adequate amounts, confer a health benefit on the host’ (10). For colic, it is postulated that probiotics act by promoting conditions in the intestinal microbiome that activate protective host factors and inhibit the pathogenic effects of bacteria (11). Studies comparing infants with and without colic have shown differences in their gut microbiota, suggesting that IC may be associated with an atypical colonization pattern and/or lack of microbial diversity (12). Interestingly, preliminary data also suggest that probiotics may have a role in neurodevelopment and mood in children and adults, representing a ‘microbiome-gut-brain axis’ (13–15). Based on this evidence, there has been increasing interest among researchers, clinicians, and parents in exploring the role of probiotics in the management of IC.

OBJECTIVE

This brief review aims to collect and synthesize the evidence from published randomized controlled trials (RCTs) and systematic reviews (SRs) of RCTs on the effectiveness of probiotics in the treatment and prevention of IC.

SEARCH METHODS

We performed comprehensive database searches in MEDLINE, EMBASE, CINAHL, Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews for studies published from January 1, 2000 to July 11, 2018. We chose this time frame because 1) we wanted to ensure that the most recent literature was informing our review and 2) probiotics are a relatively new topic in the literature and most of the relevant studies would have been performed in recent years. The search strategy is described in Appendix 3 and was limited to children and the English language. To limit our Medline search to retrieve only RCTs, we used a Cochrane highly sensitive search strategy for identifying randomized trials (16). In addition to searching the databases, we conducted citation searches (whereby studies that cited our included RCTs were also screened for inclusion), hand searches in Google Scholar and screened reference lists of relevant articles.

Furthermore, we summarized the most recent systematic review and compared this with the results of our search.

SELECTION CRITERIA FOR RCTS

Studies were included if they investigated the treatment or prevention of IC in full-term infants, were randomized by any method, included a control group, used any live probiotic, had a minimum of 15 participants in each group, were peer-reviewed and published in English (17). We excluded studies on primarily preterm infants and studies that looked exclusively at prenatal use of probiotics. For treatment trials, we included studies if they used Wessel’s or modified Wessel’s criteria to diagnose colic, and used a structured diary to record daily crying time as one of the measured outcomes (4). For prevention trials, we included studies if they used daily crying time or a surrogate outcome for crying/fussing, such as use of medications for colic and physician visits. We summarized SRs if they reviewed RCTs of any live probiotics for infantile colic.

DATA COLLECTION AND ANALYSIS

The process of study selection is presented using a PRISMA (Preferred Reporting Items for Systematic Review and Meta-analysis) flow diagram (18). Data extraction for study characteristics and critical appraisal of RCTs was performed by one author (LF) and verified by another (MK). The quality of evidence was evaluated using the risk of bias tool from the Cochrane Handbook on Systematic Reviews of Interventions and Review Manager (16,19) and, where sufficient data were available, the GRADE (Grading of Recommendations, Assessment, Development and Evaluations) criteria (20). While the risk of bias tool evaluates the quality of methods in each study, GRADE assesses the quality of evidence on the effectiveness of an intervention for a specific clinical outcome across studies, based on five dimensions: study design, risk of bias, inconsistency, indirectness, and imprecision (20,21). There are four possible GRADE scores: high (at least four points), moderate (three points), low (two points), and very low (one or less). Any reported adverse events of using probiotics in infants were collected from the reviewed studies and from recent SRs/publications about the safety of probiotics in children.

FINDINGS

The database searches yielded 155 reports, and two additional reports were found by screening reference lists of relevant articles. There were 15 studies (total population n=1,495) that met our inclusion criteria (see Appendix 2 for PRISMA flow diagram): 1) nine RCTs of treatment (n=570) and 2) six RCTs of prevention of IC (n=925). Below, we summarize published SRs and describe the findings of our search.

Treatment of infantile colic

Findings of previous SRs

We found four SRs (1,2,22,23) that assessed seven different RCTs on the use of probiotics for treatment of IC. Six trials (11,24–28) met our inclusion criteria; the seventh involved only nine infants, i.e., did not have minimum 30 participants (29). Five (11,24–27) of the six included RCTs used the probiotic Lactobacillus reuteri DSM 17938 (L reuteri). Meta-analyses performed by all three SRs showed that L reuteri benefitted breast-fed term infants compared to placebo. The most recent one (23) used individual patient data and found breast-fed infants receiving L reuteri averaged less crying/fussing at all time points (day 21 adjusted mean difference −46.4 minutes/day) and were two to three times as likely as breast-fed infants in the placebo group to experience treatment success, defined as >50% reduction in crying time (day 21 adjusted incidence ratio 2.07); differences in these outcomes between groups were noted by day 7. For breast-fed infants, number needed to treat (NNT) was 2.6. No such differences were shown in formula-fed infants.

Our updated summary

In addition to the six RCTs above (11,24–28), we found three small subsequent trials (30–32) (Table 1) yielding a total of 570 infants in eight RCTS. Seven of the nine RCTs reported reduced crying times with use of probiotics, of which five trials used L reuteri (n=315) (11,24,25,27,32), one trial used Lactobacillus rhamnosus GG (LGG) (n=30) (30), and one trial used a synbiotic (seven probiotics plus a prebiotic) (n=45) (31). The remaining two trials showed no benefit of probiotics on crying times (25,28). The first, by Dupont et al. (n=53) (28) used a formula enriched with alpha-lactalbumin plus LGG and Bifidobacterium infantis and found no difference in crying times, although irritability and agitation decreased significantly (P=0.036). In the second, Sung et al. in Australia published the largest trial to date (n=127), including a significant proportion of formula-fed infants (59%), and showed increased crying/fussing (mean +49 minutes, P=0.02) at 1 month of age in the L reuteri group (P=0.02) (25).

Table 1.

Probiotics for treatment of infantile colic- study characteristics and results

First author (year)	Sample size, age, feeding mode	Probiotic intervention- dose, control (duration)	Outcomes (measurement unit)	Results in intervention/ control group	Statistical test results
Savino (2018)	N=54, term infants <12 weeks, breast-fed	L. reuteri – once/day, placebo (30 days)	Daily crying/fussing minutes, day 30 (mean)	Int: 74.67 (SD=25.04) Ctl: 147.85  (SD= 37.99)	(95% CI: −87.32; −59.15) (P=0.001)
			Treatment success‡, day 30	Int: 75% Ctl: 36.4%	Odds ratio:0.19; RR: 039; RRR: 0.61; NNT: 2.56
Partty (2015)	N=30, term infants <6 weeks, Breast-/ formula-fed	L. rhamnosus GG- once/day, placebo (28 days)	Daily crying minutes reported by parents, (mean)	Int: 70 (SD=36) Ctl: 138 (SD=108)	(P=0.05*)
Chau (2015)	N=52, term infants 3–24 weeks, breast-fed	L. reuteri – once/day, placebo (21 days)	Daily crying minutes, (median)	Int: 60 (IQR=35–99) Ctl: 102 (IQR=61–148)	Difference = −42 (−74 to −10)*, P=0.045
			Total crying minutes over 21 days, (mean)	Int: 1,719 (SD=750) Ctl: 2,195 (SD=764)	Difference = −477 (95% CI: −53 to −900)*, (P=0.28)
			Treatment success‡, day 21 (%)	Int: 71% Ctl: 21%	RR=(3.3; 95% CI:1.55–7.03)*
Kianifar (2014)	N=45, term infants 2–16 weeks, breast-fed	Mixed probiotics¶ once/day, placebo (30 days)	Daily crying minutes, (mean)	Int: 28.8 (SD=9.7) Ctl: 63.5 (SD=10)	(P<0.001*)
			Treatment success‡, day 28 (%)	Int: 87% Ctl: 46%	NNT: (2.5), (P<0.01*)
Sung (2014)	N=127, term infants <13 weeks, 59% formula-fed at entry	L. reuteri – once/day, placebo (30 days)	Daily crying/fussing minutes, day 30 (mean)	Int: 229 (SD=137) Ctl: 191 (SD=103)	Difference: (49; 95% CI:8 to 90)*, (P=0.02*) (in favour of control group)
			Treatment success‡, day 30 (%)	Int: 40% Ctl: 48%	Odds ratio: (0.7; 95% CI: 0.4 to 1.3, P=0.23)
Szajewska (2013)	N=80, term infants <5 months, breast-fed	L. reuteri – once/day, placebo (21 days)	Daily crying minutes (median)	Int: 75 (IQR=60–90) Ctl: 128 (IQR=116–150)	Difference = −53 (−83 to −45), (P<0.0001*)
			Treatment success‡, day 21 (%)	Int: 98% Ctl: 38%	RR=1.6 (95% CI:1.3–2.1)* NNT: 3 (95% CI: 2–5)*
Dupont (2010)	N=53, 3–12 weeks, formula-fed	Formula with mix probiotics¥, placebo (30 days)	Daily crying	Did not differ between groups	NS
			Irritability and agitation as secondary outcome (NS)	Int: −39.8% Ctl: −13.6%	(P=0.036*)
Savino (2010)	N=46, 2–16 weeks, breast-fed and mothers dairy-free	L. reuteri – once/day, placebo (21 days)	Reduction of daily crying time <3 h/day, (median)	Int: 35 (IQR=85) Ctl: 90 (IQR=148)	(P=0.022*)
			Treatment success, day 21 (%)	Int: 96% Ctl: 71%	(P=0.036*)
Savino (2007)	N=83, 3 weeks to 3 months, breast-fed	L. reuteri – once/day, placebo (28 days)	Reduction of the daily average crying time <3 h/day, (median)	Int: 51 (Range = 26–105) Ctl: 145 (Range = 70–191),	Difference: –94 (–102 to –76)*, (P<0.001*)

*Statistically significant; Int: Intervention group; Ctl: control group; SD: statdard deviation; IQR: inter quartile range; RR: relative risk; RRR: relative risk reduction; NNT: number needed to treat; NS: not stated; ^‡Treatment success is the number/percentage of infants that showed reduction of daily crying time ≥50%; MD: mean difference (mean intervention – mean control); RR: relative risk (probiotics/placebo); ^¶1-billion cfu of L. casei, L. rhamnosus, Streptococcus hermophiles, Bifidobacterium breve, L. acidophilus, B. infantis, L. bulgaricus, and FOS (fructooligosaccharide); ^¥Alpha-lactalbumin enriched with probiotics (10⁷ cfu each of Lactobacillus rhamnosus and Bifidobacterium infantis).

Quality of evidence

The risk of bias assessment (16) showed that seven of the nine RCTs (11,24–26,30–32) have a low risk of bias for the main components including sequence generation, concealment of allocation, blinding, incomplete data addressed, and selective outcome reporting. Two studies had high risk of bias: one (27) did not fulfill the concealment of allocation and was not double blinded and the other (28) did not report sequence generation, concealment of allocation, and blinding. Seven RCTs reported industry support (e.g., probiotics were provided), although four indicated that the supporting firms did not have any role in the study design, analysis, or manuscript preparation (Appendix 1, part 1).

Data on treatment success of L reuteri was available for assessment by GRADE (21), yielding scores of 3+ for breast-fed infants (moderate quality of evidence) and 2+ (low quality) when the data included both breast- and formula-fed infants (Table 2). Other probiotics and outcomes, including crying times, could not be assessed due to insufficient data.

Table 2.

GRADE analysis of probiotic Lactobacillus reuteri in infantile colic*

Outcome (number of studies)	Population (number of participant)	Relative effect (95% CI)	Anticipated absolute effects (95% CI)			Certainty	Comments
					Difference
Treatment success† (five RCTs)1–5	Breast (four trials) and formula-fed (1 trial) (n=365)	RR 1.68 (0.97–2.90)	41.2%	69.3% (40–100.0)	28% more (1.2 fewer to 78.4 more)	⨁⨁◯◯ LOWa,b	High heterogeneity between RCTs (I2=83%)
Treatment success† (four RCTs)1–4	Breast-fed only (n=238)	RR 2.15 (1.13–4.09)	37.6%	80.9% (42.5–100.0)	43.2% more (4.9 more to 116.2 more)	⨁⨁⨁◯ MODERATEa	High heterogeneity between RCTs (I2=79%)
GRADE working Group; grades of evidence
⨁⨁⨁⨁ High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
⨁⨁⨁◯ Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
⨁⨁◯◯ Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
⨁◯◯◯ Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1- Savino (2018), 2- Chau (2015), 3- Szajewska (2013), 4- Savino (2010), 5- Sung (2014).

^aInconsistency in the results

^bOne RCT did not show significant difference

The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI Confidence interval; RR Risk ratio.

*Generated using GRADEpro GDT software adopted from GRADEpro Guideline Development Tool (21).

†Treatment success was defined as >50% reduction in crying/fussing time from baseline.

Prevention of infantile colic

Findings of previous SR

One SR (1) included seven RCTs (33–39) (n=1,461) assessing probiotics for prevention of infantile colic, of which three did not meet our inclusion criteria: two included pregnant mothers (33,34) and one involved only preterm newborns (39). Due to insufficient evidence, this SR did not support probiotics use for prevention of infantile colic.

Our updated summary

Our search identified six RCTs, of which four (35–38) (n=352, all formula-fed infants) were included in the SR above (1) and two were published since (n=573): one with breast-fed infants (40) and one with breast- and formula-fed infants (41). Five different probiotic formulations were studied in the six trials, given daily for 28 to 180 days to healthy term newborns. Only two RCTs showed reductions in crying times (38,41). Two trials reported significant prophylactic effects of L reuteri for reduction in use of medications (cimetropium bromide and simethicone) and in number of physician encounters (40,41) (Table 3).

Table 3.

Probiotics for preventing infantile colic study characteristics and results

First author (year)	Sample size, age, feeding mode	Probiotic intervention-dose, control (duration)	Outcome (measurement unit)	Results in intervention/ control group	Statistical test results
Savino (2015)	N=105, term infants <10 days, breast-fed	L. reuteri+Vit D3 once/day, Vit D3 (84 days)	Medication use‡, (Number of infants received the medications)	Cimetropium: Int: 1 Ctl: 24 Simethicone: Int: 11 Ctl: 48	Cimetropium: RR=0.04 (95% CI: 0.01–0.31)* (P<0.0001)*, NNT: 2.35 Simethicone: RR=0.24 (95% CI: 0.14–0.41)* (P<0.0001)*; NNT: 1.49
Indrio (2014)	N=468, term infants <1 week, Breast-/formula-fed	L. reuteri - once/day, placebo (90 days)	Daily crying minutes, (mean)	Int: 37.7 (SD=33.8) Ctl: 70.9 (SD=51.9)	P<0.01*
Roze (2012)	N=88, term infants <3 days, formula-fed	a-lactalbumin/symbiotic infant formula¶, placebo (30 days)	Daily crying minutes, (mean)	Int: 75.6 (SD=22.3) Ctl: 89.3 (SD=33.2)	P=0.03*
Vlieger (2009)	N=100, term infants <1 week, formula-fed	Formula with 2 probiotics£- once/day, placebo formula (90 days)	Daily crying hours, (mean)	Int: 1.8 (SD=1.0) Ctl: 1.8 (SD=0.9)	P>0.05
Weizman (2006)	N=59, term infants 3–65 days, formula-fed	Two formulas† once/day, Control (28 days)	Daily crying hours, (mean)	Int A†: 1.9 (SD=0.6) Int B: 1.7 (SD=0.4) Ctl: 1.9 (SD=0.5)	P=0.581
Vendt (2006)	N=105, term infants 0–60 days, Formula-fed	Formula with LGG¥, placebo (180 days)	Daily crying hours, (mean)	Int: 6.2 (SD=1.8) Ctl: 6.1 (SD=1.4)	P>0.05

*Statistically significant; SD: standard deviation; RR: relative risk (probiotics/placebo); NNT: number needed to treat; ‡Cimetropium bromide use at least three times/week or Simethicone use at least five times/week; ¶Lactobacillus rhamnosus LCS-742 (1.4 × 10⁸ cfu/100 mL) plus Bifidobacterium longum infantis M63 (1.4 × 10⁸ cfu/100 mL) plus prebiotics (95% galacto-ologosaccharides and 4% short-chain fructo-oligosaccharides); †Intervention-1: formula with 10⁷ cfu L. reuteri ATCC 55730 and intervention-2: formula with Bifidobacterium lactis (BB-12); £10⁷ cfu Bifidobacterium animalis (Bb-12) and 10⁷ cfu L. paracasei ssp. Paracasei; ¥L. rhamnosus GG (LGG).

Quality of evidence

Risk of bias in five of six RCTs (35–38,41) was low. The sixth RCT (40) did not conceal the allocation and was not double blinded. All six RCTs received industry support; two studies indicated that the supporting companies did not have any role in the study nor in manuscript preparation (40,41) (Appendix 1, part 2).

Due to significant heterogeneity of the trials, including types and formulations of probiotics, patient populations, and outcome measures, GRADE was not assessed.

SAFETY OF PROBIOTICS

Thirteen (11,24–28,30–32,35–37,41) of 15 RCTs in our review required participants (total n=1,221) to record adverse events (AEs) and two did not (38,40); no AEs related to use of probiotics were reported. Additionally, a pilot study on critically ill children admitted to intensive care found no adverse events occurred in patients given Lactobacillus casei Shirota. There was no evidence of colonization or bacteremia in samples obtained from study participants (42). A recent systematic review (43) of all available information on the safety of probiotics in humans and animal models concluded that the overwhelming existing evidence suggests that serious adverse events related to probiotics are exceedingly rare; populations at high risk of invasive infections were generally critically ill, postoperative/hospitalized, or immunocompromised. Consistent with these findings, a population-based evaluation on the safety of probiotics found that increased consumption of Lactobacillus species in Finland did not increase serious harms (44).

LIMITATIONS

There are important gaps in the literature that limit the generalizability of our results, including the limited study of formula-fed infants in the treatment trials, heterogeneous probiotic regimens, and the potential influence of diet (e.g., dairy-free maternal diet, experimental formulas), medications (e.g. proton-pump inhibitors), and geography (e.g., Australia, Canada, Europe) on outcomes. Methodological limitations, such as small sample sizes (most studies had <100 participants), and substantial heterogeneity, including the species and formulation of probiotics, inclusion criteria (e.g., family history of atopy, maternal elimination diet), and study design affected the quality of evidence and interpretation of our results.

IMPLICATIONS

The results of our rapid review are consistent with recommendations by the American Academy of Pediatrics and the Canadian Paediatric Society, which state that there is insufficient evidence to recommend routine use of probiotics in the management of IC (45,46). However, moderate-quality evidence of some benefits is emerging: in colicky full-term breast-fed infants who are otherwise healthy, L reuteri DSM 17938 appears to reduce crying time by approximately 46 minutes per day. Studies have used a standard dose of 1 × 10⁸ CFU once daily, given as five drops on a spoon or administered directly into the infant’s mouth, with onset of effect noted by 7 days of treatment and lasting at least 21 days. While probiotics have been shown to be very safe, vulnerable populations require close medical supervision, such as children who are immunocompromised or chronically/seriously ill.

There is insufficient evidence to guide recommendations for treatment of formula-fed infants with colic, or for the use of probiotics other than L reuteri for treating infant colic. There is also insufficient evidence for daily use of probiotics in healthy neonates for prevention of infantile colic (47).

CONCLUSION

This rapid review identified 4 systematic reviews and 15 randomized controlled trials reporting on the effectiveness of probiotics for the treatment and prevention of IC in healthy term infants. There is limited but favourable evidence supporting a trial of probiotics (specifically, L reuteri DSM 17938) to reduce crying due to colic in breast-fed infants. While the routine use of probiotics for the management of IC cannot yet be recommended, some clinicians and families may be interested in a therapeutic trial, given the favourable risk: benefit ratio. Evidence is insufficient to guide recommendations for formula-fed infants, for use of other probiotics, and for prevention of colic.

Funding Information: There are no funders to report for this submission.

Currently Held Funding for SV unrelated to this submission: University of Calgary, University of Alberta, Campus Alberta Outcomes and Public Health, Alberta Health Services, The Dianne and Irving Kipnes Foundation, Anonymous Donor, Canadian Institutes of Health Research, Women & Children’s Health Research Institute, Public Health Agency of Canada & Nova Scotia Health Research Foundation, Lotte & John Hecht Memorial Foundation

Potential Conflicts of Interest: All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

Ethics approval: No ethics approval was required for this work.

Appendix 1. Assessment of risk of bias in the studies of probiotics in infantile colic*.

Appendix 2. PRISMA diagram of probiotics in infantile colic evidence summary*.

Appendix 3. probiotics.