The Efficacy and Safety of the Probiotic Bacterium Lactobacillus reuteri DSM 17938 for Infantile Colic: A Meta-Analysis of Randomized Controlled Trials

Abstract

Objective: To evaluate the efficacy and safety of Lactobacillus reuteri DSM 17938 for treating infantile colic.

Methods: 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.

Results: 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.

Conclusions: 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.

Fig 1. Flowchart of studies included in the meta-analysis.

Fig 2. Forest plot comparing treatment effectiveness between L. reuteri group and placebo group under per-protocol analysis.

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.

Fig 3. Forest plot comparing treatment effectiveness between L. reuteri group and placebo group under intention-to-treatment analysis.

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.

Fig 4. Sensitivity analyses of treatment effectiveness between the L. reuteri group and the placebo group under intention-to-treatment analysis.

One week (a), two weeks (b), three weeks (c) and four weeks (d).

Fig 5. Forest plot comparing crying time between the L. reuteri group and the placebo group.

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.

Fig 6. Sensitivity analyses comparing crying time between the L. reuteri group and the placebo group.

One week (a), two weeks (b), three weeks (c) and four weeks (d).

Fig 7. Forest plot comparing growth parameters between the L. reuteri group and the placebo group on weight, length and head circumference.

The statistical method was Cohen’s d, the effect measure was WMD, and the analysis method was the fixed-effects model.