A meta-analysis of probiotic efficacy for gastrointestinal diseases

Abstract

Background: Meta-analyses on the effects of probiotics on specific gastrointestinal diseases have generally shown positive effects on disease prevention and treatment; however, the relative efficacy of probiotic use for treatment and prevention across different gastrointestinal diseases, with differing etiology and mechanisms of action, has not been addressed.

Methods/principal findings: We included randomized controlled trials in humans that used a specified probiotic in the treatment or prevention of Pouchitis, Infectious diarrhea, Irritable Bowel Syndrome, Helicobacter pylori, Clostridium difficile Disease, Antibiotic Associated Diarrhea, Traveler's Diarrhea, or Necrotizing Enterocolitis. Random effects models were used to evaluate efficacy as pooled relative risks across the eight diseases as well as across probiotic species, single vs. multiple species, patient ages, dosages, and length of treatment. Probiotics had a positive significant effect across all eight gastrointestinal diseases with a relative risk of 0.58 (95% (CI) 0.51-0.65). Six of the eight diseases: Pouchitis, Infectious diarrhea, Irritable Bowel Syndrome, Helicobacter pylori, Clostridium difficile Disease, and Antibiotic Associated Diarrhea, showed positive significant effects. Traveler's Diarrhea and Necrotizing Enterocolitis did not show significant effects of probiotcs. Of the 11 species and species mixtures, all showed positive significant effects except for Lactobacillus acidophilus, Lactobacillus plantarum, and Bifidobacterium infantis. Across all diseases and probiotic species, positive significant effects of probiotics were observed for all age groups, single vs. multiple species, and treatment lengths.

Conclusions/significance: Probiotics are generally beneficial in treatment and prevention of gastrointestinal diseases. Efficacy was not observed for Traveler's Diarrhea or Necrotizing Enterocolitis or for the probiotic species L. acidophilus, L. plantarum, and B. infantis. When choosing to use probiotics in the treatment or prevention of gastrointestinal disease, the type of disease and probiotic species (strain) are the most important factors to take into consideration.

Figure 1. PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) flow diagram showing an overview of the study selection process.

Figure 2. The effect size (risk ratio) for the overall effects of probiotics in the prevention and treatment of gastrointestinal (GI) diseases including the 95% confidence intervals.

The diseases: Antibiotic associated diarrhea (AAD), Clostridium difficile disease (CDD), Helicobacter pylori positive (HPP), Irritable bowel syndrome (IBS), Infectious diarrhea (ID), Necrotizing Enterocolitis (NEC), Traveller's diarrhea (TD), and Pouchitis are labelled as well as the mean effect sizes for each disease. The author, date, measure (risk ratio (95% CI), and p value are shown. Risk ratios below one favor the probiotic while risk ratios above one favor the placebo.

Figure 3. The effect size (risk ratio) for gastrointenstinal diseases and for probiotic species.

(A) The effect size including the 95% confidence intervals for the total events of Antibiotic associated diarrhea (AAD), Clostridium difficile disease (CDD), Helicobacter pylori positive (HPP), Irritable bowel syndrome (IBS), Infectious diarrhea (ID), Necrotizing Enterocolitis (NE), Traveller's diarrhea (TD), and Pouchitis during which probiotics were taken. (B) The effect size including 95% confidence intervals for the type of probiotic species that were used to treat and prevent gastrointestinal disease. The species that were used were VSL#3, Lactobacillus rhamnosus GG (LGG), Saccromyces boulardii, Bifidobacterium infantis, Lactobacillus acidophilus, Lactobacillus casei, Clostridium butyricum, Enterococcus faecum, Lactobacillus plantarium, Bifidobacterium lactis and Lactobacillus acidophilus combined with Bifidobacterium infantis. Risk ratios below one favor the probiotic while risk ratios above one favor the placebo.

Figure 4. The effect size (risk ratio) for the subgroup analyses on age groups, dose, treatment length and multiple or single probiotic species.

(A) The effect size including the 95% confidence intervals for the age groups that had taken the probiotic vs. the controls. Age groups included were: adults (>18 yrs), children (3≤18 yrs) and infants (0–3 yrs). (B) The effect size including the 95% confidence intervals for dose of probiotic. The doses that were included were: 1–9×10¹¹, 10¹² CFU/day; 1–5.5×10⁶, 10⁷, 10⁸ CFU/day; 1–9×10⁹ CFU/day; 1–5×10¹⁰ CFU/day. (C) The effect size including the 95% confidence intervals for treatment length. Treatment lengths that were included were: 1–2 weeks, 3–4 weeks, 5–8 weeks and 9–240 weeks. (D) The effect size including the 95% confidence intervals for multiple or single species of probiotics. Probiotics that contain more than one species were considered multiple species, while probiotics only administered as one species were considered single species. Risk ratios below one favor the probiotic while risk ratios above one favor the placebo.

Figure 5. Funnel plot asymmetry used to determine publication bias.

Log of the risk ratios were plotted against the standard error of the risk ratio of each study to identify asymmetry in the distribution of trials. Gaps in the funnel plot suggest potential publication bias. The synthesis estimate and the 0.01 limit are shown to distinguish asymmetry.