Effects of Probiotics in Adults with Gastroenteritis: A Systematic Review and Meta-Analysis of Clinical Trials

1. Introduction

Gastroenteritis poses serious public health concerns in both high- and low-income countries. It is one of the leading causes of morbidity and mortality worldwide [1]. Globally, the estimated annual cost of healthcare and loss of productivity due to gastroenteritis is about $60 billion of which developing countries bear the highest burden [1,2]. According to the Centers for Disease Control and Prevention (CDC), gastroenteritis is the most prevalent infectious disease syndrome in the United States, accounting for over 350 million illnesses annually and about 200,000 deaths, with the elderly having higher risks of mortality [3,4]. The symptoms of gastroenteritis could range from mild asymptomatic infections to life-threatening conditions and deaths [5].

Probiotics are supplements or foods that contain live non-pathogenic microorganisms, which can maintain and improve microbial balance in the gastrointestinal tract [6]. Some beneficial effects of probiotics relevant to the treatment and prevention of gastroenteritis include: reduction of invasion and colonization of the intestine by pathogenic organisms, modification of host immune response, and reduction of pH in the intestine [6,7]. Although, some studies confirmed that probiotics have anti-inflammatory and antimicrobial effects and help maintain good bacteria in the gut, results from some other studies are inconclusive and conflicting [8]. Use of specific probiotic preparations should be based on evidence from well-designed clinical trials. Therefore, the purpose of this study was to present the results of a systematic review and meta-analysis carried out to examine the effects of probiotics in adults with acute and chronic gastroenteritis of multiple etiologies.

2. Materials and Methods

A systematic literature search was conducted from February – May, 2021 using PubMed database as the primary data source. Other research databases included were Google Scholar, MEDLINE, and Scopus. The search keywords in PubMed were “effects of probiotics in gastroenteritis”, Medical Subject Headings (MeSH) Terms were: "effect*" OR "outcomes" OR "impact" OR "efficacy*" OR "efficaciousness*" AND "Probiotics" AND "Gastroenteritis" AND "Clinical Trial".

Articles which met the following criteria were included in the review: 1) Studies published between years 1990-2022; 2) Only clinical trial study designs; 3) Studies related to the effects of probiotics in gastroenteritis; 4) Studies with partcipants between 19 – 55 years of age; 5) Full text articles; and 6) Articles written in English language. Figure 1 shows the detailed search strategies for this study using the preferred reporting items for systematic reviews and meta-analyses (PRISMA) [9].

Quality appraisal methods

Studies were appraised for quality using CADIMA [10]. CADIMA is a free tool which is available online for managing articles for systematic review including automated duplicate removal, uploading PDF articles, and documentation fo the review process. Through CADIMA, standards for critical appraisal and the rating scale were defined. We followed the critical appraisal tools for systematic reviews developed by the University of Adelaide, South Australia [10]. A rating scale from 0 to 4 was based on the following criteria: (1) Sample size: greater than 30 = 1; smaller sample = 0; (2) Randomized controlled trials = 1; not randomized, no controls = 0; (3) Studied both safety and efficacy = 1; otherwise = 0; (4) Standard and objective evaluation criteria = 1, otherwise = 0. Based on the above-mentioned criteria, we rated each of the 35 studies independently from a range of 0 to 4.

3. Results

Of the 1,865 research articles identified through database search, 1,544 articles were excluded for not being done in adults. Of the remaining 321 studies, 190 were excluded because the studies did not involve clinical trials; 84 were excluded for not having full articles available; and 12 were excluded for studies not meeting the eligibility criteria, and/or not being related to gastroenteritis. The results presented below include analyses of 35 studies for the systematic review (Figure 1).

The total sample size for the 35 studies was 4,577, ranging from 15 to 777 samples in individual study, with a median of 44. Only 12 studies (34%) had a sample size of more than 100.

3.1. Type of Illnesses Associated with Diarrhea

We found that the majority of the studies (51%, 18 of 35) focused on effectiveness of probiotics in the treatment of IBDs (ulcerative colitis and Crohn’s disease), 11% (4 of 35) of the study patients had pouchitis (inflammation that occurs in the lining of a pouch created during surgery to treat ulcerative colitis or certain other diseases), about 9% (3 of 35) had antibiotic-induced diarrhea, 6% (2 of 35) had diarrhea due to Helicobacter pylori, and one each (2.9%) had diverticulitis or acute watery diarrhea due to Vibrio cholerae and enterotoxigenic Escherichia coli infection, (Figure 2).

Detailed information about the 35 selected studies [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45], the type of gastroenteritis studied, type of probiotics used and their effectiveness are presented in Table 1.

3.2. Probiotic Strains Used and Follow-up

Most of the studies (60%, n = 21) administered multiple strains of probiotics while the remaining 14 studies administered single strain of probiotics. The most commonly administered probiotic strains were Lactobacilli, Bifidobacteria, Streptococcus, and Escherichia. In a few studies, probiotics were administered as an adjuvant therapy with another conventional treatment, such as an anti-inflammatory drug for IBDs [15,20,32], rehydration therapy for acute watery diarrhea [30], or with a combination of treatments comprising antibiotics and proton pump inhibitor (PPI) (Esomeprazole) for chronic gastritis or duodenal ulcer due to H. pylori infection [43].

Patients’ follow-up protocols varied widely, ranging from as low as 10 days to 2 years depending on the illness type. Ten studies followed patients for 12 months or more, 12 studies followed patients for 3-11 months while another 14 followed patients for 3 months or less.

3.3. Quality Appraisal Findings

Due to having no major inter-observer variations in the evaluation of the quality of the studies, an average of the 4 scores was presented in Table 3. Out of 35 studies reviewed, 19 (55%) scored high (4 out of 4), 15 (43%) scored moderate (3 out of 4), and only one was rated poor (2 out of 4). Among the 27 studies which were proven effective for the treatment of probiotics, the majority (63%, n = 14) were of high quality (score 4 out of 4), 44% (n = 12) scored moderate quality (score 3 out of 4), and only 1 (4%) scored poor rating (score 2 out of 4).

3.4. Efficacy and Safety of Probiotics

The outcome measures were considered favorable if studies reported resolution, remission, improvement or no relapse of gastroenteritis after treatment. Of the 35 studies reviewed, 27 (77%) showed a favorable response after using probiotics, 7 (20%) showed that probiotics were ineffective, and 1 study conducted in Iran [34] was inconclusive (Table 1). Probiotics were most effective in the treatment or prevention of gastroenteritis due to IBDs (Figure 2). Of the various forms of probiotics, a mixture of multiple probiotic strains, in the form of VSL #3, was found most effective in treating patients with IBDs [22,29,39,40] and drug-induced enteropathy [31,35]. However, probiotics were found ineffective in five studies in patients with ulcerative colitis or Crohn’s diseases [12,14,17,26,27], and also in two other studies – one in 20 patients with pouchitis in Finland [23], and the other study in 183 patients with a severe form of acute watery diarrhea due to V. cholerae and E. coli infections in Bangladesh [30] (Table 1).

All probiotics administered in these studies, including those which were proven ineffective, were well tolerated by patients and no adverse side effects were reported. However, several studies cautioned the use of probiotics among immunocompromised patients due to safety concerns in such patients [8,50].

3.5. Effectiveness of Probiotics, as Evaluated by Meta-Analysis of 22 Clinical Trials

Due to the unavailability of relevant data, meta-analysis was conducted using 22 out of 35 (63%) studies. Table 2 shows relative risk and 95% confidence intervals (CIs) of the effect of probiotics in each study. Risk ratios observed a protective effect in 50% of the studies (n = 11); however, 95% CI included 1 in each of them. The pooled relative risk was 0.99, with 95% CI being 0.90 and 1.09. Test for the overall effect showed a p-value of 0.37, meaning that there was not enough evidence to indicate that the intervention had a significantly more protective effect compared to controls. The value of τ² indicated low variation of true effects. The Higgins H test (I²) was 6%, indicating a homogeneous nature of the weights of the studies evaluated in the meta-analysis.

In Figure 3, the Forest Plot, risk ratio and 95% CI included the line of no effect, and the p-value for overall effect was 0.85. Because of negligible heterogeneity among the studies, we can rely on the aggregated estimate more as the majority or all individual studies reached the same conclusion.

3.6. Overall Risk of Bias by Categories of Bias

As shown in Figure 4 and Figure 5, the domains evaluated in this review were as follows: (a) Random Sequence Generation: It is a form of selection bias (biased allocation to interventions) resulting from insufficient generation of a randomized sequence; (b) Allocation Concealment: It is also a selection bias (biased allocation to interventions) owing to insufficient allocation concealment prior to assignment; c) Blinding of Participants and Personnel: This is a performance bias due to participants' and personnel's knowledge of the assigned interventions during the study; d) Blinding of Outcome Assessment: This is a detection bias due to outcome assessors' knowledge of the allocated interventions; e) Incomplete Outcome Data: This is an attrition bias due to the quantity, character, or treatment of incomplete outcome data; f) Selective Reporting: This is a bias in reporting owing to the selective reporting of outcomes; g) Other Bias: This category encompasses biases caused by issues not covered elsewhere in the table. In Figure 4 and Figure 5, the rating scales of bias were high (red), low (green), and unclear (yellow).

Based on Figure 5, only one study [36] (Steed et al. 2010) was judged having ‘high risk’ of bias for two domains: (1) Random Sequence Generation (selection bias). Reviewer’s comments: There was no description of the randomization process, and the domain selection reporting (reporting bias); (2) Incomplete Outcome Data (attrition bias). Reviewer’s comments: handling of incomplete outcome data was not described in detail.

3.7. Assessment of Publication Bias

Figure 6 shows that the larger studies cluster around the top of the plot, whilst smaller studies are spread across the bottom of the plot. This is an ideal funnel plot where the included studies have scattered either side of the overall effect line in a symmetrical manner. There is no severe asymmetry to either side, so we conclude that publication bias was not present.

4. Discussion

4.1. Effective Treatment with Probiotics

Out of 35 studies that were investigated for the systematic review, our meta-analysis included 22 (63%) studies. The overall results of systematic review and meta-analysis demonstrated a mixed effect of probiotics therapy in gastroenteritis in adults. However, studies found that either short-term or long-term administration of probiotics were safe.

Our analysis suggests that based on the systematic review per se, probiotics were an effective therapeutic alternative or an adjuvant therapy for gastroenteritis in 27 out of 35 (77%) of studies, whereas meta-analysis of the aggregated data did not provide enough evidence to support statistically significant protective effects of probiotics in the study subjects with gastroenteritis. The systematic analysis also showed that probiotics were not effective in 7 (20%) patients with gastroenteritis.

Disease conditions that improved on probiotic treatment mostly included IBDs due to ulcerative colitis and Crohn’s disease. Other disease entities were chronic gastritis, pauchitis, H. pylori infections, and drug-induced enteropathies. There were no particular disease entities that were found ineffective in the treatment with probiotics. In this review, a combined therapy of probiotics with either nonsteroidal anti-inflammatory drugs, proton pump inhibitors, or antibiotics showed effective results in the treatment of IBDs, chronic gastritis and infection due to H. pylori. Probiotics were also effective in the disease prevention or prevention of relapses in 8 studies [11,21,25,26,27,37,42,44]. Findings of this review also suggested that probiotics were useful in improving patients’ quality of life in two studies (29, 41) in addition to reducing morbidity due to chronic gastroenteritis.

Our systematic review data are consistent with earlier reports which confirmed the effectiveness of probiotics in gastroenteritis. However, the effectiveness of probiotics is dependent on dose, strains used in probiotics, duration of therapy, and the type of illnesses [7,47]. One may compare the effects of probiotics in adults with the ones described in studies in children, primarily to evaluate the effectiveness of probiotics on gastroenteritis due to different etiologies [48].

4.2. Ineffectiveness of Probiotics

It is noteworthy that approximately 20% (n = 7) of the studies reported that probiotics were not effective in treating gastroenteritis, which creates an interesting discrepancy. The reasons for an ineffective outcome of these studies are mostly speculative at this point. The study of acute watery diarrhea due to V. cholerae and E. coli infections in Bangladesh [30] were more severe in nature compared to gastroenteritis due to other causes, and these patients with acute watery diarrhea needed intravenous rehydration. In such severe form of diarrheal diseases, the effectiveness of probiotics remains a question. However, in addition to the results of using a lyophilized form of Streptococcus feacium SF68, further studies may be evaluated by using other forms of probiotics or using a combination of probiotics and antibiotics in acute diarrhea. In the other cases of failure, all happened to be in patients with IBDs. Matsuoka et al (2018) [27] emphasized that the reason of lack of effectiveness may be due to inadequate dose of probiotics administered, the route of administration, or the inability to confirm improvement by using endoscopic and/or imaging procedures. Another report having no improvement after a synbiotic therapy in patients with mild-to-moderate degree of IBDs could be attributable to a small sample size and absence of more specific and objective biomarkers of inflammation such as fecal calprotectin and histologic scores to diagnose the disease [12].

4.3. Safety Issues

Our review did not find any safety concerns of probiotics in any of the studies. However, several studies described and warned about major safety issues in using probiotics, including but not limited to, systemic infections, deleterious metabolic activities, excessive immune stimulation in susceptible individuals, gene transfer and gastrointestinal side effects [49,50]. Other studies cautioned that the administration of probiotics among vulnerable populations, especially immunocompromised individuals should be carefully considered [8,50]. However, a solution to this conundrum may lay in the idea of making the report of adverse events involving probiotics in a mandatory and standardized way, thereby improving safety of products and reliability of data [51].

4.4. Limitations of the Study

One limitation of this systematic review and meta-analysis is that it focused on limited sources of studies available for synthesis using PubMed, MEDLINE, Google Scholar, and Scopas. Our study cannot be generalized because only studies published in English language and those having free full texts were reviewed. However, the nature of a comprehensive systematic review, using 30 years of timeframe of the published reports, and inclusion of study findings of radomized controlled clinical trials only may limit possibilities of bias. In addition, risks of bias assessment were low in all studies except one in our meta-analysis. A very low value (6%) of Higgins test (I²) of heterogeneity also indicates a homogeneous nature of the studies evaluated.

5. Conclusions

Our meta-analysis provided new information in contrast to the data obtained from the systematic review on the effectiveness of probiotics in gastroenteritis in adults. Even the data gathered from the systematic review had a mixed effect of probiotics – although probiotics were effective in the treatment and reducing relapses of chronic inflammatory gastrointestinal conditions in majority of the adults (78%), the review shows that probiotics are ineffective in about 20% of the patients. More importantly, the pooled data of the meta-analysis demonstrated no statistically significant protective effects of using probiotics. However, there were paucity of data from developing countries. Further studies are needed to confirm whether probiotics can restore the gut microflora and improve gastroenteritis as a single therapy or as an adjunct therapy with other conventional treatments of the infection.