The paper, published in the Annals of Internal Medicine, warns that randomised clinical trials (RCTs) assessing the efficacy of probiotics, prebiotics, and synbiotics often fail to report potential harms and adverse event (AE) data.
Without this safety data, the team behind the study warn that “it is not possible to broadly conclude that the interventions are safe.”
“Our systematic review of 384 recently published RCTs assessing probiotics, prebiotics, or synbiotics found that harms reporting for these interventions often was missing, insufficient, or inadequate,” reported the team – led by Aïda Bafeta from the French Institute of Health and Medical Research (Inserm).
However, experts responsible for clinical research in the area have questioned the study’s conclusions, while the International Probiotic Association (IPA) also suggested the findings do not represent a wider picture - with poor guidelines and inconsistent reporting of safety data a problem across nutrition and medical literature.
Indeed, the IPA noted that the title of the paper and phrasing of its conclusions – which focus on potential ‘harm’ – are suggestive and ‘besides the truth’.
“By the use of wording like this the authors suggest that AEs did occur but were not reported,” the IPA told NutraIngredients. “This is a serious accusation.”
“The use of the term ’harm’ is also ambiguous - more correct would have been to use ‘adverse events’.”
Lack of reporting does not mean ‘unsafe’
The new study suggests that ‘harms reporting’ for probiotic, prebiotic and symbiotic interventions were almost always “missing, insufficient, or inadequate.”
Indeed, they team noted that one-third of the trials studied gave no information on harms, and just 2% were found to adequately report key safety components.
As such the authors warned that these inadequacies in reporting harms-related results may lead to an inaccurate safety profile and “erroneous decision-making, with major consequences.”
Responding to the study, experts noted that while the issue of reporting adverse events in journal publications needs to be addressed, claims that probiotics, prebiotics and synbiotics may be ‘unsafe’ are inappropriate.
“The authors may be expert in how to conduct assessment of harm, and I agree that careful assessment of side effects of human studies are warranted, but their conclusions about safety of probiotics are unsubstantiated,” commented Professor Gregor Reid from Western University and the Lawson Health Research Institute, Canada.
“Just because someone didn’t perform a safety study on eating lettuce, [it] doesn’t mean that lettuce isn’t safe to eat. Sure, someone might choke on it and die, but it’s so rare we can say that lettuce is safe to eat,” he told NutraIngredients. “For the billions of doses of probiotics administered, if there were any real suggestions of safety concerns they would have become evident by now.”
The IPA added that while there clearly seems to be room to improve the observation and reporting of AEs in pre- and probiotic intervention studies, the study makes no attempt to compare the situation of safety data reporting in probiotics with other medical or nutrition interventions.
“As such the reader is left with the suggestion that only research into probiotics and prebiotics is of poor quality and does not report AEs,” the IPA warned.
'Safety should never be presumed
While all seem to agree that the issue of ‘unreported’ safety data needs addressing, there are mixed opinions of how this can be done … and who is at fault for the current situation.
“Many researchers in this area think that a detailed evaluation of potential harms is not necessary,” noted Bafeta and colleagues. “Caution is needed, however, particularly when considering these interventions for vulnerable or critically ill persons.”
“The safety profile of an intervention should never be presumed; rather, it should be rigorously evaluated and reported,” they added. “We need to promote more active implementation and endorsement of recommendations for standardising evaluation and reporting methods, such as developing a reporting tool with specific templates.”
The IPA noted that researchers are alert for potential negative side effects, commonly classified as Adverse Events (AEs) or Serious Adverse Events (SAE).
“Many clinical trials are supervised by an independent Data Safety Monitoring Board,” the trade association noted, however it said that how any adverse events are reported in the scientific literature is an issue of ‘constant attention’.
A lack of consensus … or just boring to journals?
“The data reported in the publication, are not surprising because they have been the subject of several other recent reviews,” added the IPA. “What is surprising is that apparently there still is no consensus among medical journals about the level of detail and the format of reporting adverse effect.”
Adding to the debate, Professor Glenn Gibson of Reading University in the UK noted that pro-, pre- and synbiotics all have ‘excellent’ track records of success and safety in comparison to other microbial interventions like antibiotics, which are widely accepted and used.
He told NutraIngredients that in more than 50 human studies on people of various ages and states of health he and his team have performed, there have been no problems “aside from extremely rare gaseous feelings in a tiny minority of individuals.”
“In all cases, volunteers completed diaries on tolerance, GI symptoms, other issues, general wellbeing,” said Gibson.
“We usually summarise tolerance data in publications,” he added. “But it is also worth mentioning that some journals have asked us to remove his when there are no issues (as it is a neutral and uninteresting observation I suppose).”
However, Gibson noted that if there was any concern over safety then it would not be possible to gain ethical permission – meaning that the trial would not be able to commence in the first place.
“In our experience, we must cover this in the ethical permission application document. All potential risks are asked about and we have to respond to these,” he told us.
A route forward?
The authors of the report concluded that an evaluation of the benefit–risk balance “should always be included in RCT reports,” noting that an international and collective effort is ‘urgently needed.’
The IPA said the main takeaway from the publication is the importance of accurate reporting of AEs.
“But more than that: the methodology of clinical and laboratory investigations used to detect any AEs could and should be included in clinical studies submitted to databases like ClinicalTrials.Gov,” it said.
“Medical journals should [also] be stricter in what they demand in terms of AE reporting,” it added.
Source: Annals of Internal Medicine
Published online, ahead of print, doi: 10.7326/M18-0343
“Harms Reporting in Randomized Controlled Trials of Interventions Aimed at Modifying Microbiota: A Systematic Review”
Authors: Aïda Bafeta, et al