To evaluate the efficacy and safety of Lactobacillus reuteri DSM 17938 for treating infantile colic.
A systematic literature retrieval was carried out to obtain randomized controlled trials of L. reuteri DSM 17938 for infantile colic. Trials were performed before May 2015 and retrieved from the PubMed, EMBASE, Cochrane library, CNKI, WanFang, VIP, and CBM databases. Data extraction and quality evaluation of the trials were performed independently by two investigators. A meta-analysis was performed using STATA version 12.0.
Six randomized controlled trials of 423 infants with colic were included. Of these subjects, 213 were in the L. reuteri group, and 210 were in the placebo group. Lactobacillus reuteri increased colic treatment effectiveness at two weeks (RR = 2.84; 95% CI: 1.24–6.50; p = 0.014) and three weeks (relative risk [RR] = 2.33; 95% CI: 1.38–3.93; P = 0.002) but not at four weeks (RR = 1.41; 95% CI: 0.52–3.82; P = 0.498). Lactobacillus reuteri decreased crying time (min/d) at two weeks (weighted mean difference [WMD] = –42.89; 95% CI: –60.50 to –25.29; P = 0.000) and three weeks (WMD = –45.83; 95% CI: –59.45 to –32.21; P = 0.000). In addition, L. reuteri did not influence infants’ weight, length or head circumference and was not associated with serious adverse events.
Lactobacillus reuteri possibly increased the effectiveness of treatment for infantile colic and decreased crying time at two to three weeks without causing adverse events. However, these protective roles are usurped by gradual physiological improvements. The study is limited by the heterogeneity of the trials and should be considered with caution. Higher quality, multicenter randomized controlled trials with larger samples are needed.
Citation: Xu M, Wang J, Wang N, Sun F, Wang L, Liu X-H (2015) The Efficacy and Safety of the Probiotic Bacterium Lactobacillus reuteri DSM 17938 for Infantile Colic: A Meta-Analysis of Randomized Controlled Trials. PLoS ONE 10(10): e0141445. https://doi.org/10.1371/journal.pone.0141445
Editor: Vincenzo Lionetti, Scuola Superiore Sant'Anna, ITALY
Received: July 6, 2015; Accepted: October 7, 2015; Published: October 28, 2015
Copyright: © 2015 Xu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Infantile colic, often described as excessive crying, is prevalent in the first three months of life, in approximately 20% of infants . Because infantile colic usually begins at approximately two weeks of age and improves by month four, previous reports have described it as a benign and self-limiting condition, albeit one that often puts stress on the parents’ mental health . In addition, several observational studies have reported infantile colic to be associated with childhood migraine [3, 4] and attention deficit and hyperactivity disorder . Therefore, an attitude change toward infantile colic is necessary. In fact, even though it is not long lasting, infantile colic should be treated proactively.
The treatment of infantile colic consists of dietary, pharmacological and behavioral interventions and complementary and alternative therapies . Simethicone, a safe, over-the counter drug for decreasing intraluminal gas, has been promoted as an agent to decrease colicky episodes in the 1990s , anticholinergic drugs, dicyclomine, herbal teas and cimetropium, have also been used to treat infantile colic, but they did not show obvious improvement in the treated group. Other treatments, such as: placing colicky infants in car-ride simulators or near a clothes dryer or vacuum cleaner  also did not reach treatment goal. Besides these treatment protocols, the use of probiotics to treat infantile colic was recently proposed. Several beneficial effects of probiotics on the host intestinal mucosal defenses system have been identified. These include blocking pathogenic bacterial effects by producing bacteriocidal substances and competing with pathogens and toxins for adherence to the intestinal epithelium [8, 9]. Two reviews from 2013 have concluded that the probiotic Lactobacillus reuteri (L. reuteri) DSM 17938 is promising for treating breastfed infants with colic [10, 11]. Although previous results indicated that L. reuteri was an effective treatment strategy for excessive crying in exclusively breastfed infants with colic, only three small studies found this to be true, and these did not offer sufficient evidence to support probiotics’ use in managing manage colic or preventing excessive infant crying. With the emergence of similar research studies, various results have become available on this topic. In 2014, Sung and colleagues  concluded that L. reuteri was ineffective in both breastfed and formula-fed infants with colic.
Thus, to comprehensively evaluate the efficacy and safety of L. reuteri (especially DSM 17938) for treating infantile colic, we performed a global search of published randomized controlled trials (RCTs) on this topic. We then applied quantitative analyses with the goal of providing evidence for clinical decision-making.
We included subjects described in six RCTs performed prior to May 2015. Subjects consisted of infants aged three to six months with term delivery who suffered from colic. All infants were exclusively or predominantly breastfed and formula fed. All the infants enrolled in the studies needed to be consecutively recruited from general pediatrician practices and outpatient populations. The diagnosis of infantile colic accorded with Wessel’s criteria: crying or fussy episodes lasting three or more hours per day and occurring at least three days in the 1 week prior to enrolment. Infantile parents were instructed to maintain and record the daily crying (in minutes) according to the Barr diaries  or not. Included infants could not have gastrointestinal disorders, acute or chronic illness, or use of any antibiotic products within seven days prior to the study. There is no limitation on whether infants have allergy to milk protein or a family history of allergy. All analyses in this meta-analysis were based on previous published studies; thus, no ethical approval or patient consent was required.
Infants were assigned at random to receive L. reuteri DSM 17938 or placebo. The primary end points were treatment effectiveness (defined as percentage of children achieving a ≥ 50% reduction in daily average crying time) and duration of crying (min/d). The secondary end points were infantile growth parameters, maternal mental health (measured with the Edinburgh Postnatal Depression Scale or questionnaire) and adverse events.
We searched PubMed (1966–2015.5), EMBASE (1974–2015.5) and the Cochrane library (Issue 4 of 12, April 2015) using a searching strategy that combined MeSH/Emtree terms and free text words: Excessive crying, Infantile Colic*, Colic [Mesh/Emtree], Colic, child*, bab*, infant*, kid*, pedia*, pedo*, Lactobacillus reuteri, L. reuteri and Lactobacillus reuteri [Mesh/Emtree]. We also searched the following databases in Chinese: CNKI, CBM, WanFang, and VIP. Retrieval dates came from time of database creation to May 2015. To avoid missed terms, we also searched articles using Google Scholar. In addition, we manually checked the references listed at the end of studies to locate potentially eligible research studies.
Data Extraction and Quality Evaluation
Two investigators independently read the titles, abstracts, and full texts using the following steps: (1) examining titles and abstracts to remove obviously irrelevant studies, (2) retrieving the full texts of potentially relevant trials, (3) examining the full texts for compliance with eligibility criteria, and (4) making final decisions on study inclusion and proceeding to data collection. From the included studies, the investigators extracted baseline information on infant subjects (e.g., treatment strategy, dose, and duration) and detailed methods used in the study design (e.g., publication year, study settings, designs, methods of randomization, allocation concealment, blinding). Disagreements were resolved by discussing them with a third investigator. Some data about crying time was described as median, we translated it into mean to aggregate depending on experience , when n≤15, estimated mean , estimated standard deviation , median (m), low and high end of the range (a and b, respectively); when 15<n≤25, estimated mean , estimated standard deviation ; when 25<n≤70, estimated mean , estimated standard deviation ; when 70<n, estimated mean , estimated standard deviation .
Each study was independently assessed for its methodological quality by the previous investigator. The criteria were based on those described in the Cochrane Reviewers’ Handbook 5.1.0. They included selection bias, performance bias, detection bias, attrition bias, and reporting bias domains, random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting and other biases.
The meta-analysis was performed using STATA version 12.0 (STATA Corporation, College Station, TX, USA). Relative risk (RR) and weighted mean difference (WMD) were chosen as effective sizes for dichotomous and continuous variables, which were described with a 95% confidence interval (CI). Before analysis, we calculated statistical heterogeneity using I2 statistics and assigned low heterogeneity, moderate heterogeneity, and high heterogeneity I2 values of 25%–50%, 50%–75% and > 75%, respectively . If the heterogeneity of studies was low (P > 0.1, I2 < 50%), we adopted a fixed-effects model for quantitative analysis. If the heterogeneity was moderate or high (P < 0.1, I2 > 50%), possible sources of heterogeneity were analyzed, and subgroup analysis against confounding factors should be divided, after which a random effects model was applied . In addition, investigators conducted sensitivity analyses to explore possible explanations for heterogeneity on the overall pooled estimates. Sensitivity analysis was calculated by omitting a single study in each turn. A statistically significant difference for their conclusion was P < 0.05.
Process for Selecting Trials
As shown in Fig 1, 165 potentially relevant studies were identified and screened for retrieval. Twenty-three studies were excluded because of duplications, and 129 studies were excluded after reading their titles and abstracts. Among the remaining 13 studies, three were conference abstracts, two were reviews and one was a RCT comparing L. reuteri DSM 17938 with simethicone, and therefore excluded. Following this step, seven RCTs were assessed for eligibility. After reading, one RCT that included infantile diseases other than colic was excluded. Finally, six RCTs [12, 17–21] were included in this meta-analysis.
Characteristics of Included Trials and Quality Evaluation
The main characteristics of the trials included in our meta-analysis are shown in Table 1. The number of infants in RCTs varied from 29 to 167. A total of 423 infants with colic were included in the meta-analysis; of these, 213 were in the L. reuteri group, and 210 were in the placebo group. All infants were born at term, adequate for gestational age, and aged less than five months at recruitment. Infants in five RCTs were exclusively or predominantly breastfed; only in the Sung study  were they breastfed and formula fed.
In six RCTs, the oral dose of L. reuteri was five drops (0.2 × 108 colony-forming units [CFU] per drop) once daily. The duration of treatment differed among studies. Four RCTs [17, 18, 20, 21] treated infants for 21 days, and two RCTs [12, 19] treated them for one month. Active study products in RCTs consisted of a suspension of freeze-dried L. reuteri DSM 17938 in an oil mixture, whereas the placebos contained the same ingredients but without the live bacteria.
As shown in Table 2, all six RCTs reported random sequence generation from computerized randomization. Allocation concealment was detailed in three studies [12, 17, 21]. Five RCTs [12, 17–19, 21] reported a double-blind design, and one study  reported a single-blind design. Intention-to-treat analysis was adopted in three trials [12, 17, 18]. Thus, we determined that these six trials were high quality.
Primary End Point
Five studies [12, 17, 18, 20, 21] compared the treatment effectiveness of L. reuteri and placebo. As shown in Fig 2, the aggregated results of these studies were divided into four subgroups according to time point. Heterogeneity in one-, two-, three- and four-week subgroups were I2 = 0% (P = 0.445), I2 = 81.5% (P = 0.001), I2 = 82.5% (P = 0.000) and I2 = 87.7% (P = 0.000), respectively. We adopted a random effects model (using per-protocol analysis), which suggested that L. reuteri supplementation possibly increased the rate of treatment effectiveness at one week (RR = 2.43; 95% CI: 1.41–4.16; P = 0.001), two weeks (RR = 3.43; 95% CI: 1.30–9.01; P = 0.012) and three weeks (RR = 2.42; 95% CI: 1.35–4.35; P = 0.003) but not at four weeks (RR = 1.69; 95% CI: 0.82–3.48; P = 0.158).
The statistical method used was the Mantel–Haenszel (M-H) method, the effect measure was RR, and the analysis method was the random effects model.
We used intention-to-treatment analysis to aggregate the data. As shown in Fig 3, heterogeneity in the one-, two-, three- and four-week subgroups were I2 = 4.0% (P = 0.373), I2 = 80.1% (P = 0.002), I2 = 81.1% (P = 0.000) and I2 = 94.8% (P = 0.000), respectively. We adopted a random effects model, and under intention-to-treatment analysis, data showed the same results as in the previous per-protocol analysis: L. reuteri treatment was able to increase the rate of treatment effectiveness at one week (RR = 2.03; 95% CI: 1.20–3.43; P = 0.008), two weeks (RR = 2.84; 95% CI: 1.24–6.50; P = 0.014) and three weeks (RR = 2.33; 95% CI: 1.38–3.93; P = 0.002) but not at four weeks (RR = 1.41; 95% CI: 0.52–3.82; P = 0.498). Meanwhile, the Number Needed to Treat (NNT) is a good epidemiological measure for evaluating the efficacy of L. reuteri. After calculation (NNT = 1/risk difference [RD]), the NNT of treatment effectiveness was 2.56 (1/0.39, RD was 0.39) at two weeks and 2.23 (1/0.44, RD was 0.44) at three weeks.
The statistical method used was the Mantel–Haenszel method (M-H), the effect measure was RR, and the analysis method was the random effects model.
We further performed sensitivity analyses to explore stability. As shown in Fig 4, treatment effectiveness at one, two, three and four weeks changed to varying degrees after removing any one RCT. The studies with the greatest influence on overall results were Savino et al.  and Chau et al. . Treatment effectiveness at one week was 2.92 RR (95% CI: 0.68–12.58; P = 0.151) after removing Savino et al.  and 1.99 RR (95% CI: 0.88–4.50; P = 0.098) after removing Chau et al. . These results differed significantly from those of the intention-to-treat analysis.
Five studies [12, 17, 18, 20, 21] reported crying time in the L. reuteri group and the placebo group. Data in three studies [12, 17, 21] was translated into mean±standard deviation. As shown in Fig 5, the aggregated results of these studies were divided into four subgroups according to treatment time. The heterogeneity in one-, two-, three- and four-week subgroups were I2 = 82.1% (P = 0.000), I2 = 72.9% (P = 0.005), I2 = 57.1% (P = 0.053) and I2 = 94.3% (P = 0.000), respectively. Thus, we adopted the random effects model for aggregating results. The results demonstrated that L. reuteri was able to decrease crying time (min/d) at one week (WMD = –28.37; 95% CI: –49.25 to –7.49; P = 0.008), two weeks (WMD = –42.89; 95% CI: –60.50 to –25.29; P = 0.000), three weeks (WMD = –45.83; 95% CI: –59.45 to –32.21; P = 0.000) and four weeks (WMD = –56.32; 95% CI: –89.49 to –23.16; P = 0.001).
The statistical method used was Cohen’s d (it is an effect size used to indicate the standardised difference between two means, also widely used in meta-analysis), the effect measure was WMD, and the analysis method was the random effect model.
We further performed sensitivity analyses. As shown in Fig 6, crying time at one, two, three and four weeks changed to varying degrees after removing any one RCT. The studies with the greatest influence on the overall results were Szajewska et al.  and Mi et al. . Crying time at one week had a WMD of –27.60 (95% CI: –56.17 to 0.97; P = 0.058) after removing Mi et al. ; crying time at four weeks had a WMD of –46.81 (95% CI: –130.65 to 37.03; P = 0.274) after removing Szajewska et al.  and a WMD of –36.78 (95% CI: –100.43 to 26.87; P = 0.257) after removing Mi et al. . These results differed significantly from those of the intention-to-treat analysis.
One week (a), two weeks (b), three weeks (c) and four weeks (d).
Based on the above data, the results showed that L. reuteri possibly was able to increase the rate of treatment effectiveness for infantile colic at two and three weeks and to decrease crying time at two and three weeks.
Secondary End Points
We also compared the growth parameters of both groups based on the two RCTs [18, 21] that reported such data (Fig 7). The results suggested that L. reuteri supplementation for 21 days to one month did not affect the growth rate of infants: heterogeneities in weight, length and head circumference were I2 = 0.0% (P = 0.829), I2 = 0.0% (P = 0.705), and I2 = 0.0% (P = 0.679), respectively. After we adopted the fixed-effects model, we found that L. reuteri use did not influence infant weight (WMD = –0.13; 95% CI: –0.43 to 0.17; P = 0.385), length (WMD = –0.11; 95% CI: –1.14 to 0.93; P = 0.842) or head circumference (WMD = 0.50; 95% CI: –0.17 to 1.16; P = 0.143).
The statistical method was Cohen’s d, the effect measure was WMD, and the analysis method was the fixed-effects model.
It was necessary to assess maternal mental health to reflect the parents’ psychological burden. Only two RCTs [12, 21] reported this information using the Edinburgh Postnatal Depression Scale after the infant received treatment. The results in Mi et al.  indicated that L. reuteri was able to decrease the Edinburgh Postnatal Depression Scale at one to four weeks. Sung et al.  reported that this score was similar in the L. reuteri group and the placebo group at one month.
Adverse events were reported in five RCTs [12, 17, 18, 20, 21]. Savino et al.  reported rhinitis in the L. reuteri group (n = 1) and eczema, fever, otalgy, and gastroesophageal reflux in the placebo group. Four studies [12, 17, 20, 21] reported no adverse events associated with the probiotic therapy or the placebo.
L. reuteri DSM 17938 is a new probiotic strain that does not carry potentially transferable traits for tetracycline and lincomycin resistance . L. reuteri is used to treat gastroenteritis in children [23, 24]. The potential mechanism is from the anti-inflammatory effects, because there are number of publications showing anti-inflammatory effects of this probiotic in neonatal animals . Inflammation has been also described in infants with colic , and recent evidence suggests that probiotics might offer some benefits to infants with colic as well.
In our meta-analysis, six RCTs contributed to relatively stable results: L. reuteri DSM 17938 increased the rate of treatment effectiveness at one, two and three weeks, but not at four weeks. L. reuteri also decreased crying time at one, two, three and four weeks. Sensitivity analyses demonstrated that L. reuteri increased the rate of treatment effectiveness at two and three weeks and decreased crying time at two and three weeks. Specially, after sensitivity analyses, Savino et al.  and Chau et al.  in results of treatment effectiveness, Szajewska et al.  and Mi et al.  in results of crying time at different time points were been found that these studies influenced the stability of results, we drew a comprehensive conclusion based on above analyses, so results from individual study at different time did not affect the final conclusions. Thus, treatment effectiveness and decreasing crying time, two primary outcomes, indicate that L. reuteri could play a positive role in the treatment of infantile colic, particularly at two to three weeks.
It is important to note that two possible reasons explain why L. reuteri does not demonstrate a protective role at one and four weeks. First, not only is L. reuteri not found in every individual, but it is rarely found. Rhoads et al. has studied more than 35 infants with colic and found only 2/34 patients had lactobacilli in the stool before probiotic treatment, and they were both L. gasseri . Instead, dietary supplementation may be necessary to introduce lactobacilli and to effectively colonize the intestine of healthy infants. After oral intake, colonization of L. reuteri begins rapidly within days of ingestion before it plays a positive role in the digestive system . Therefore, at one week, the treatment effectiveness and crying times in the L. reuteri group were equivalent to those in the placebo group. Second, the mean age of the two study groups was 28.5–52.5 days in the L. reuteri group and 28.6–48.3 days in the placebo group. Therefore, after treatment lasting 21 days or one month, infants were beginning to improve physiologically, and the role of L. reuteri was not as important to increasing rate of treatment effectiveness and reduction of crying time. This fade-out effect probably explained why there was no difference between the two groups at four weeks. In Sung et al. , the mean age of the two groups was greater than that in the other studies: 52.5 days for the probiotic group and 48.3 days for the placebo group, 33.5 days and 35.2 days for the probiotic and placebo group in other five studies. Thus, mean age of the babies was 19 days and 13.1 days greater than other five studies in the probiotic and placebo group respectively. This difference of the studies could explain why Sung et al.  was the only one that demonstrated negative results for L. reuteri treatment. Based on the two points above, we conjecture that L. reuteri is effective for treating infantile colic if a favorable therapeutic window is chosen.
No serious adverse events were observed in the six studies, and there were no differences in growth parameters between the two groups in our meta-analysis. Even up to the maximum tested dosage of 1010 CFU of L. reuteri per day, no significant differences in standard medical laboratory tests were found . Four studies [12, 17, 20, 21] reported no adverse events associated with the probiotic therapy or the placebo. This indicates that these studies did not prospectively evaluate for the presence of side effects (gastrointestinal, upper respiratory, etc.). Without studies documenting that patients were given a diary card, adverse event frequency cannot be ascertained. This is important because live microorganisms can be linked to side effects such as sepsis and even death in vulnerable infants . Thus, adverse events associated with the probiotic therapy or the placebo should be marked and raised in the next study. Maternal mental health was an indirect indicator of the efficacy differences between the two groups, but only two studies reported this.
Our meta-analysis has several potential limitations that should be taken into account. First, mode of feeding is associated with the occurrence of infantile colic. In this study, infants in five RCTs [17–21] were exclusively or predominantly breastfed; only in Sung et al.  were infants breastfed and formula fed. Thus, we could not analyze the effect of feeding mode by subgroup. Second, in most studies, infant crying time recorded by parents varied widely and was based on subjective parental judgments. Third, some data about crying time is described as median, we translate it into mean to aggregate depending on experience. Fourth, three studies [12, 17, 20] reported the information of family history of allergy, but some previous trials indicate that L. reuteri may have an effect on the immune system and allergic symptoms in children , which may have contributed to the differences in the findings. Fifth, Roos et al.  focused more on the faecal samples from colicky infants treated with L. reuteri, rather than the clinical indicator. This study only reported number of responders without documenting crying time, so its result of "improvement" may be more subjective than the other five studies.
In conclusion, the beneficial effects of L. reuteri DSM 17938 might be demonstrated in infantile colic. The probiotic possibly increases treatment effectiveness and decreases crying time at two to three weeks, without causing adverse events. However, these protective roles are usurped by physiological improvements over time. The conclusions of this study were limited by heterogeneity of the included RCTs and need to be considered with caution. Higher quality, multicenter RCTs with larger patient samples are needed.
Conceived and designed the experiments: XHL. Performed the experiments: MX JW. Analyzed the data: NW FS LW. Wrote the paper: MX JW.
- 1. Savino F, Tarasco V. New treatments for infant colic. Curr Opin Pediatr. 2010; 22: 791–797. pmid:20859207
- 2. Savino F, Ceratto S, De Marco A, Cordero di Montezemolo L. Looking for new treatments of Infantile Colic. Ital J Pediatr. 2014; 40: 53. pmid:24898541
- 3. Sillanpaa M, Saarinen M. Infantile colic associated with childhood migraine: A prospective cohort study. Cephalalgia. 2015; pmid:25754178
- 4. Romanello S, Spiri D, Marcuzzi E, Zanin A, Boizeau P, Riviere S, et al. Association between childhood migraine and history of infantile colic. JAMA. 2013; 309: 1607–1612. pmid:23592105
- 5. Kaymaz N, Uzun ME, Cevizci S, Yildirim S, Ilcin M, Topalogolu N, et al. Attention deficit and hyperactivity disorder and infantile colic. Minerva Pediatr. 2015; 67: 391–399 pmid:25923450
- 6. Metcalf TJ, Irons TG, Sher LD, Young PC. Simethicone in the treatment of infant colic: a randomized, placebo-controlled, multicenter trial. Pediatrics. 1994; 94: 29–34 pmid:8008533
- 7. Roberts DM, Ostapchuk M, O'Brien JG. Infantile colic. Am Fam Physician. 2004; 70: 735–740 pmid:15338787
- 8. Yan F, Polk DB. Probiotics and immune health. Current opinion in gastroenterology. 2011; 27: 496–501. pmid:21897224
- 9. Vitetta L, Manuel R, Zhou JY, Linnane AW, Hall S, Coulson S. The overarching influence of the gut microbiome on end-organ function: the role of live probiotic cultures. Pharmaceuticals (Basel). 2014; 7: 954–989. pmid:25244509
- 10. Sung V, Collett S, de Gooyer T, Hiscock H, Tang M, Wake M. Probiotics to prevent or treat excessive infant crying: systematic review and meta-analysis. JAMA pediatrics. 2013; 167: 1150–1157. pmid:24100440
- 11. Anabrees J, Indrio F, Paes B, AlFaleh K. Probiotics for infantile colic: a systematic review. BMC Pediatr. 2013; 13: 186. pmid:24238101
- 12. Sung V, Hiscock H, Tang ML, Mensah FK, Nation ML, Satzke C, et al. Treating infant colic with the probiotic Lactobacillus reuteri: double blind, placebo controlled randomised trial. BMJ. 2014; 348: g2107. pmid:24690625
- 13. Barr RG, Kramer MS, Boisjoly C, McVey-White L, Pless IB. Parental diary of infant cry and fuss behaviour. Arch Dis Child. 1988; 63: 380–387 pmid:3365007
- 14. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol. 2005; 5: 13. pmid:15840177
- 15. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–560. pmid:12958120
- 16. Armitage P, Berry G, Matthews JNS. Analysing Means and Proportions, Stat Methods Med Res. 4th ed. Blackwell Science Ltd; 2008
- 17. Szajewska H, Gyrczuk E, Horvath A. Lactobacillus reuteri DSM 17938 for the management of infantile colic in breastfed infants: a randomized, double-blind, placebo-controlled trial. J Pediatr. 2013; 162: 257–262. pmid:22981952
- 18. Savino F, Cordisco L, Tarasco V, Palumeri E, Calabrese R, Oggero R, et al. Lactobacillus reuteri DSM 17938 in infantile colic: a randomized, double-blind, placebo-controlled trial. Pediatrics. 2010; 126: e526–533. pmid:20713478
- 19. Roos S, Dicksved J, Tarasco V, Locatelli E, Ricceri F, Grandin U, et al. 454 pyrosequencing analysis on faecal samples from a randomized DBPC trial of colicky infants treated with Lactobacillus reuteri DSM 17938. PloS One. 2013; 8: e56710. pmid:23468874
- 20. Mi GL, Zhao L, Qiao DD, Kang WQ, Tang MQ, Xu JK. Effectiveness of Lactobacillus reuteri in infantile colic and colicky induced maternal depression: a prospective single blind randomized trial. Antonie Van Leeuwenhoek. 2015; 107:1547–1553. pmid:25876529
- 21. Chau K, Lau E, Greenberg S, Jacobson S, Yazdani-Brojeni P, Verma N, et al. Probiotics for infantile colic: a randomized, double-blind, placebo-controlled trial investigating Lactobacillus reuteri DSM 17938. J Pediatr. 2015; 166: 74–78. pmid:25444531
- 22. Rosander A, Connolly E, Roos S. Removal of antibiotic resistance plasmids from Lactobacillus reuteri ATCC 55730 and characterization of the resulting daughter strain, L reuteri DSM 17 938. Appl Environ Microbiol. 2008; 74: 6032–6040. pmid:18689509
- 23. Szajewska H, Urbanska M, Chmielewska A, Weizman Z, Shamir R. Meta-analysis: Lactobacillus reuteri strain DSM 17938 (and the original strain ATCC 55730) for treating acute gastroenteritis in children. Beneficial microbes. 2014; 5: 285–293. pmid:24463209
- 24. Dinleyici EC, Group PS, Vandenplas Y. Lactobacillus reuteri DSM 17938 effectively reduces the duration of acute diarrhoea in hospitalised children. Acta Paediatr. 2014; 103: e300–305. pmid:24579935
- 25. Liu Y, Fatheree NY, Dingle BM, Tran DQ, Rhoads JM. Lactobacillus reuteri DSM 17938 changes the frequency of Foxp3+ regulatory T cells in the intestine and mesenteric lymph node in experimental necrotizing enterocolitis. PloS One. 2013; 8: e56547. pmid:23437165
- 26. Rhoads JM, Fatheree NY, Norori J, Liu Y, Lucke JF, Tyson JE, et al. Altered fecal microflora and increased fecal calprotectin in infants with colic. J Pediatr. 2009; 155: 823–828.e1. pmid:19628216
- 27. Wolf BW, Garleb KA, Ataya DG, Casas IA. Safety and Tolerance of Lactobacillus reuteri in Healthy Adult Male Subjects. Microb Ecol Health Dis. 1995; 8: 41–50.
- 28. Doron S, Snydman DR. Risk and safety of probiotics. Clin Infect Dis. 2015; 60 Suppl 2: S129–134. pmid:25922398
- 29. Abrahamsson TR, Jakobsson T, Bottcher MF, Fredrikson M, Jenmalm MC, Bjorksten B, et al. Probiotics in prevention of IgE-associated eczema: a double-blind, randomized, placebo-controlled trial. J Allergy Clin Immunol. 2007; 119: 1174–1180. pmid:17349686