The results of our meta-analysis of RCTs show that patients with IBD exposed to probiotics experience side effects more frequently than those exposed to placebo.
There exists a robust body of literature on probiotics in patients with IBD; however, the majority of available studies are clinical trials designed to evaluate the efficacy of probiotics and prebiotics versus placebo, or other conventional therapies, for the achievement or maintenance of remission in patients with CD or UC
[22]. In spite of that, the safety profile of probiotics and prebiotics in patients with IBD remains less explored, probably because they have been consumed as food for hundreds of years, especially in dairy products such as yogurt (the most famous are the Sardinian one: gioddu, the Caucasian: chefir; the Russian: kumis and the Egyptian: leben and for this reason perceived as safe. Described since ancient time in the Bible as a precious food, the acid milk regained popularity with the Nobel prize I.I. Mechnikov, who suggested its consumption against senility, up to nowadays with several probiotic strains, alone or combined, commercially available in hundreds of different products. Although the widespread use of probiotics, prebiotics and synbiotics, studies usually miss to report adequately adverse events related to their use, despite in some subgroups of patients their use was associated with severe side effects
[20][21].
These findings are difficult to explain, since they are in contrast with the intrinsic meaning of probiotic “pro bios” in favor of life. For example, in a recent study conducted in a series of 200 IBD patients including both CD and UC, we observed a 93% reduction in the need for systemic steroids, hospitalization, and surgery related to the disease in CD patients, and a 100% reduction in UC patients taking probiotics for more than 75% of disease duration
[17]. Similarly, in a study including 170 IBD patients exposed to probiotics use, we noticed a reduction in the occurrence of skin lesions
[24]; however, this reduction was dependent on the amount of probiotics taken
[17][24]. Probably, the duration of treatment with probiotics may influence outcomes. In the two previous studies probiotics were taken for years, whereas in the RCTs included in our meta-analysis they were taken for a maximum of one year
[25][26]. In general, it is an increasingly accepted concept that probiotic benefits may be dose-dependent, only manifesting themselves upon achievement of a threshold dosage
[27]. Probiotic amounts in currently used formulations show a wide variability, ranging from 10
7 to 10
11 CFU/g, which likely implies a high variability in the number of viable cells included in the products. It has been ascertained that the percentage of live cells capable of driving an effective change in fecal microbiota can vary from 1% to 92%
[27]. We can assume that only formulations with a high bacterial load may exert a positive effect, while those with a low bacterial content have no effect or are even contrary to expectation. For example, daily doses of Lactobacilli equal to or greater than 10
10 CFU induced a significant reduction in the duration of diarrhea in children. For lower doses of probiotics, an increase in the duration of diarrhea was paradoxically observed
[28]. In addition to the duration and dose, the probiotic strain may play a central role in health and disease. For instance, in a double-blind placebo-controlled trial, Mangalat et al. observed a prominent proinflammatory triggering following the administration of
L. reuteri in healthy adults as revealed by increased fecal calprotectin
[29]. Moreover, some individuals taking probiotic may temporarily experience an increase in gas production and swelling, in addition to constipation, which in most cases disappears in a few weeks
[30][31]. Several lactic bacteria produce bioactive substances such as histamine, tyramine, and phenylethylamine, which may induce headaches and other complaints
[32]. These bioactive molecules are normally inactivated by mono-amino oxidases (MAO) in the intestinal wall and liver
[33]. Minderhoud reported an alteration in neuroendocrine cells in IBD patients with IBS-like symptoms associated with lower MAO activity and an increase in biogenic amines
[34]. A number of additional biogenic amines have also been isolated from bacterial strains commonly used in probiotic preparations
[35]. These observations may provide a partial explanation for the occurrence of gastrointestinal side effects in patients exposed to probiotics including abdominal cramping, nausea, soft stools, flatulence, and taste disturbance, occurring in subjects receiving probiotics
[20]. The gut microbiota resides almost completely in the colon and, to a lesser extent, in the small bowel. For this reason, it is plausible that probiotics affects especially colon microorganism communities. For the same reason treatment of IBD patients with probiotics is more effective in UC than in CD. In a review addressing the quantitative risk–benefit analysis of probiotic use in IBS and IBD, Bennet reported gastrointestinal symptoms as the most frequent side effects
[36], although it is difficult to find a clear cut between gastrointestinal symptoms generated by the natural course of IBD and those generated by probiotic exposure. Another critical point is that in RCTs comparing probiotics with placebo in IBD patients, the conventional treatment was not similar in both arms, and in some series, the difference was statistically significant
[15][26][37][38]. Lack of uniformity in the concomitant medications in each of these studies may be one limitation of this systematic review and meta-analysis. In addition, the majority of RCTs lacked standardized methods for the assessment of side effects (for instance, a validated questionnaire), or their graduation. Moreover, the strain, amount, schedule, and duration of probiotics/synbiotic treatment used in RCTs were extremely variable. Finally, but no less important, we were unable to relate a specific strain to side effect occurrence, due to the paucity of data.