Statistics from Altmetric.com
Maria is a 1.5-month-old healthy baby born at 39 weeks, mainly breastfed, although she has been fed artificial formula on a few occasions. In the last 2 weeks, she has been brought to the clinic with episodes of intense crying which last from 3 to 5 hours/day and occur 4–5 times a week. Her parents are finding it hard to soothe and calm her during these episodes, and find it very distressing. She undergoes a normal clinical examination and there do not seem to be any other associated conditions. As such, infantile colic seems to be the most probable cause of her current distress. Her parents have read that certain probiotics might help her, but you are very uncertain and want to know more.
Structured clinical question
In infantile, colic (patient) does routine administration of any strain of probiotics (intervention) effectively decreases the intensity or duration of these episodes of excessive crying (outcome) without increasing the risks of sepsis or diarrhoea?
We searched Ovid MEDLINE via PubMed, EMBASE and NHS evidence (up to December 2018 week 3). The advanced search mode was used including the terms infantile colic OR infant colic OR excessively crying OR fuss behaviour OR feeding problems OR sleeping problems OR infant* colic* AND probiotics OR Lactobacillus OR Bifidobacterium OR Saccharomyces. This search yielded 28 results. The results were limited to humans. Only English or European-language publications were examined.
These included the Cochrane Database of Systematic Reviews (current issue) in the Cochrane Library, Tripdatabase, Dialnet and Scopus. They provided one additional reference.
These articles were independently screened by two authors (MR and AD), identifying three relevant papers. All references were screened, identifying one further article.
The details of the included studies are provided in table 1: (summary of search results).
Infantile colic affects ~20% of babies worldwide.1 Although different therapies have been tested in the empirical clinical practice (eg, medicinal herbs, acupuncture, massage, hydrolysed or isolated soy-protein formulas and so on), whenever rigorous trials have been undertaken, almost none of them have shown to be more effective than placebo.2 3 The only intervention that seems to have shown certain benefits and under certain specific circumstances is Lactobacillus reuteri (LR) strain Deutsche Sammlung von Mikroorganismen (DSM) 17 938.
Although the cause of infantile colic remains unclear, it has been hypothesised that a dysbiosis (microbial imbalance) could play an important role in its aetiology.4 However, we are still far from understanding the microbiome (bacteria, viruses and yeasts with whom we keep a close symbiosis). Although probiotics have been described to form complex ecosystems with synergies and antagonisms, there are still many facts to be determined such as how these strains interact, if their immune effect might be long-lasting, whether they might have associated resistance to antibiotic, how essential fatty acids might be altered and even the biological reason that lies behind their mutagenic capacity.
Although the natural course of infantile colic is self-limited and benign, it has been postulated that the intestinal–cerebral axis could act as a key modulator for brain development in the long term, and that perhaps the alleged microbial dysbiosis might be associated with a higher risk of developing subsequent recurrent abdominal pain, allergic disease and therefore impacting brain development.4 Pärtty et al 5 randomised control trial (RCT), with a 13-year follow-up, suggested that there might be an interconnection between the intestinal microbiota in early stages of life and the subsequent development of psychological disorders, though this conclusion is significantly weakened by the number of drop outs.
These uncertainties are reflected in the 2014 European and 2017 Infectious Diseases Society of America guidelines6 7 which suggest that administration of effective probiotic strains ‘may be considered’ to reduce symptoms' severity in infantile colic. The European Society for Paediatric Gastroenterology Hepatology and Nutrition (2014 guideline6 and the ESPGHAN Working Group on probiotics/prebiotics8 recommend that LR DSM 17938 (at >1010 colony-forming units (CFU)/day, for 21 days) may be considered in infantile colic.
Table 1 shows the current evidence with this probiotic obtained from high-quality clinical trials: (1) a recent individual participant data meta-analysis (MA)9, includes data of four RCTs, and (2) two subsequent RCTs.10 11 Sung et al 9 is not a usual aggregate data MA, but rather an individual participant data MA. This approach pools raw data from individual studies with the aim of having sufficient statistical power to produce more reliable estimates of treatment effects, as well as subgroup analyses. In certain circumstances, they may have some advantages over the usual MAs. The authors of the four RCTs took part in its drafting.12–15 All of them used LR DSM 17 938 versus placebo using same dosage (1×108 CFU orally/day, during 21 days) in a total sample of 345 participants. This MA excluded three RCT due to not being double-blinded: Mi et al,16 Ashraf et al 17 and Martinelli et al.18
The review9 showed positive effects in infants receiving exclusively or predominantly (>50%) breast milk, whereas a lack of effect was registered in infants exclusively fed with formula 14, with a clinically relevant magnitude of effect. Therefore, intervention effects were clearly registered among breastfed infants, but were not statistically significant in formula-fed infants. In breastfed infants (exclusively or not), this MA found that 21 days after starting the treatment, there was a decrease in crying time of about 25 min when compared with the placebo group (adjusted mean difference: −25.4 min (95% CI −47.3 to −3.5)). When treatment success (defined as 50% reduction in the duration of crying and/or fussing) was analysed, a time-response gradient was detected at 7, 14 and 21 days, with a lower effect on 7 days (number needed to treat (NNT) 8) and a great one at 21 days (NNT 4). The probiotic-treated group was almost twice as likely as the placebo group to experience treatment success (day 21 adjusted incidence ratio 1.7 (95% CI 1.4 to 2.2)).
In Sung et al 9 MA, the risk of bias was evaluated according to the Cochrane Collaboration methods. A bias that could limit the magnitude of effect confidence was the fact that crying, and fussing were subjectively recorded in non-validated instruments in two of the RCTs.12 13 In addition, in two RCTs,14 15 caregivers recorded their estimated perception of crying and fussing time at the end of the day, potentially leading to a memory or adherence bias.
Two other RCTs have been published after Sung et al 9 MA, comparing the same probiotic (LR DSM 17938) versus placebo in exclusive or predominant breast-fed babies. In USA, Fathree et al 10 and Savino et al 11 in Italy evaluated decreased crying time, fussing and various surrogate markers, such as inflammatory, immune and microbiome variables. The treatment dosage and duration in both RCTs were higher than those of previous RCTs (5×108 CFU/day), and the treatment lasted 30 days11 and 42 days, respectively.10 The mean crying time was measured in both RCTs10 11 by using a 24-hour Barr diary (a validated scale).19 The smaller Fathree et al 10 RCT was underpowered but showed no effect in the change in the crying time. The number of responders (≥50% reduction in crying plus fussy time from baseline) at day 21 was 66% in both groups: RR=0.93 (95% CI 0.50 to 1.73).
The other RCT 11 included 52 patients and showed that infants treated with LR had a significant decrease in crying duration at the end of the study (day 30), with NNT of 3 in order to reduce 50% of the registered crying (95% CI 2 to 5).
This is currently a field of research that continuously evolves, as there are nine ongoing RCT protocols related to LR in infantile colic registered in clinicalTrial.gov.
Work on understanding which infants may benefit is also in evolution. Zmora et al 20 carried out a metagenomic study in which they concluded that the presence of probiotic in stools might not be an ideal marker to detect the probiotic incorporation in the individual microbiota. In addition, they also describe that colonisation patterns might be specific for the intestinal transcriptome of each individual. This finding, if true, might mean limiting the universal recommendation of probiotics for infantile colic, as an idiosyncratic response might be expected.
No adverse effects associated with the LR DSM 17938 strain used in healthy infants have been described so far. However, we should still remain cautious, as probiotics have been classified as functional foods by the Food and Drug Administration and European Medicines Agency, meaning that they do not need to follow the regulated guidelines that regular drugs strictly undergo, and may reflect under-reporting.
Based on these findings, LR (dosage 108 CFU/day for 21 days) might be beneficial to certain patients (such as healthy term babies with adequate birth weight and predominantly breastfed) presenting with moderate to severe infantile colic. We find evidence for a greater benefit in one RCT, which used doses fivefold higher than previous trials, in the same kind of infants.
No adverse events have been described so far in this subgroup of patients. In those particular cases where probiotics might be indicated—and since they are classified as functional food products—appropriate commercialised strains tested in high-quality RCTs should be used to ensure their safety and viability.
Clinical bottom lines
Lactobacillus reuteri (DSM 17938) decreased the average crying time in predominantly breasted babies with infantile colic (grade A).
The magnitude of clinical benefit is uncertain as based on only small studies, and might be influenced by the dosage as well as the length of administration (grade D).
The authors thank Ms Andrea Cervera Alepuz for assistance with the English translation.
Contributors MR and AB conceived the study, developed the search strategy, contributed to data collection, abstraction and interpretation. MR drafted the first manuscript. ADI and PC contributed to data collection, interpretation and provided critical revisions to the manuscript. All authors approved the final version of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.
Patient consent for publication Not required.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.