Strain-Specificity and Disease-Specificity of Probiotic Efficacy: A Systematic Review and Meta-Analysis

Abstract

Background: As the use and diversity of probiotic products expands, the choice of an appropriate type of probiotic is challenging for both medical care professionals and the public alike. Two vital factors in choosing the appropriate probiotic are often ignored, namely, the probiotic strain-specificity and disease-specificity for efficacy. Reviews and meta-analyses often pool together different types of probiotics, resulting in misleading conclusions of efficacy.

Methods: A systematic review of the literature (1970-2017) assessing strain-specific and disease-specific probiotic efficacy was conducted. Trials were included for probiotics with an identifiable strain (either single strain or mixtures of strains) that had at least two randomized, controlled trials for each type of disease indication. The goal was to determine if probiotic strains have strain and/or disease-specific efficacy.

Results: We included 228 trials and found evidence for both strain specificity and disease specificity for the efficacy of specific probiotic strains. Significant efficacy evidence was found for 7 (70%) of probiotic strain(s) among four preventive indications and 11 (65%) probiotic strain(s) among five treatment indications. Strain-specific efficacy for preventing adult antibiotic-associated diarrhea was clearly demonstrated within the Lactobacillus species [e.g., by the mixture of Lactobacillus acidophilus CL1285, Lactobacillus casei LBC80R, and Lactobacillus rhamnosus CLR2 (Bio-K+^®), by L. casei DN114001 (Actimel^®) and by Lactobacillus reuteri 55730], while other Lactobacillus strains did not show efficacy. Significant disease-specific variations in efficacy was demonstrated by L. rhamnosus GG and Saccharomyces boulardii CNCM I-745, as well as other probiotic strains.

Conclusion: Strong evidence was found supporting the hypothesis that the efficacy of probiotics is both strain-specific and disease-specific. Clinical guidelines and meta-analyses need to recognize the importance of reporting outcomes by both specific strain(s) of probiotics and the type of disease. The clinical relevance of these findings indicates that health-care providers need to take these two factors into consideration when recommending the appropriate probiotic for their patient.

Keywords: Clostridium difficile; Lactobacillus; Saccharomyces; antibiotic-associated diarrhea; disease specificity; meta-analysis; pooling data; probiotic strains; strain specificity.

Figure 1

PRISMA flow diagram of evaluated studies for randomized controlled trials for probiotic strain and disease-specificity, searched from inception of databases to February 2017.

Figure 2

Forest plot of probiotic strain-specificity. Meta-analysis of probiotic strain-specificity for six different Lactobacillus probiotics for the prevention of antibiotic-associated diarrhea in adults. Meta-analysis of 22 randomized, controlled trials by sub-group of probiotic type. Abbreviations: acid, acidophilus; CI, confidence interval; helv, helveticus; ID, identification; L, Lactobacillus; rhamn, rhamnosus; RR, relative risk.

Figure 3

Forest plot of disease-specificity. Meta-analysis of disease-specificity shown for the prevention of six different types of diseases by one strain of probiotic, Lactobacillus rhamnosus GG. A meta-analysis of 24 randomized, controlled trials in adults and children. Abbreviations: AAD, antibiotic-associated diarrhea; C, Clostridium; ID, identification; CI, confidence interval; RR, relative risk.

Figure 4

Meta-analysis of nine randomized controlled trials of three different lactobacilli-containing probiotics for the prevention of antibiotic-associated diarrhea in children and adults, sub-grouped by identical strains of Lactobacillus. Modified from the one “lactobacilli” group presented in Hempel et al. (64). Abbreviations: acid, acidophilus; CI, confidence interval; L, Lactobacillus; helv, helveticus; ID, identification; RR, relative risk.

Figure 5

Example of forest plot from an appropriate meta-analysis using probiotic strain subgroup analysis for the prevention of Clostridium difficile infections modified from Lau and Chamberlain (66). Grouping by the same probiotic strain determined that Lactobacillus rhamnosus GG was ineffective, S. boulardii was effective, and only two mixtures were significantly effective (72). Abbreviations: ID, identification; CI, confidence interval; RR, relative risk; S. boulardii, Saccharomyces boulardii CNCM I-745; L., Lactobacillus; L acid + Bifid bifidum, Lactobacillus acidophilus and Bifidobacterium bifidum strains not reported; L acid + L. casei + L. rhamn, L. acidophilus CL1285 + L. casei LBC80R + L. rhamnosus CLR2, Bio-K+^®.