This report must be framed in the context of the community knowledge initiative initiated to identify possible emerging risks associated with the consumption of food supplements other than vitamins and minerals.
The initiative aims to gather market information, including alerts from (nutri)vigilance and monitoring systems, to identify substances of concern and support the prioritization of those that may warrant action under Article 8 of Regulation (EC) No 1925/2006. This provision allows the European Commission to restrict, prohibit or place under scrutiny substances added to foods where potential health risks are identified.
The scientific report states that even if the work did not highlight “clear emerging risks”, it resulted in a shortlist of substances that could be further considered for assessment under article 8. These shortlisted substances include Coleus (Plectranthus barbatus), guggul (Commiphora mukul), toothed club moss (Huperzia serrata), black radish (Raphanus sativus L.), ginseng (Panax ginseng) and the genus Bifidobacterium.
Understanding the methodology
To reach this conclusion and provide a clear understanding of the situation, the authors followed a defined methodology.
First, case reports of adverse events were collected, and those considered relevant (i.e. related to food supplements, sufficiently documented and showing at least a likely causal relationship between consumption and outcome) were classified according to severity by applying the French Nutrivigilance system. This system grades adverse events across four levels—from 1 (mild) to 4 (lethal outcome), with Level 2 and 3 representing intermediate and severe effects, respectively. This newly published report considers only cases classified as Levels 3 and 4.
A literature search is then performed to identify further “ingredient-adverse effect” connections. Retrieved articles are assessed against predefined inclusion and exclusion criteria and assigned to one of these three scoring tiers: B0 (not documented); B1 (poorly documented, comprehending isolated clinical cases not supported by pathophysiological data, or animal studies); and B2 (well documented, including clinical or epidemiological studies or cases supported by pathophysiological data). The recently published analysis focuses specifically on the B1-rated ingredients, which are considered potential emerging risks.
Finally, those ingredients within the B1 category are ranked based on three criteria: the number of related cases, the number of related publications and the existence of human data.
Considering the vigilance signals collected in Belgium, Denmark, France, Italy, Netherlands and Portugal, ingredients ranked B1 and with a high score of concern in this report, together with the reasons for that concern, were:
Some of these ingredients with highest Score (5 or 6) are already subject to different stages of scrutiny, restriction or prohibition. The main outcome of this exercise was therefore to flag the six ingredients mentioned at the beginning, including Bifidobacterium, which received a score of 6 for infectious endocarditis based on the following criteria:
- A single case report, which adds 0 points in the ranking methodology
- More than three publications, contributing 4 points
- The availability of human data, contributing 2 points
At this point, the association between Bifidobacterium and infectious endocarditis may seem surprising. To interpret the finding correctly, it is important to note that the system uses the term “infectious endocarditis” as a synonym for bacterial infection, which may be misleading. These events are classified as Level 3 grade (severe effects). The scientific report cites six documents that contributed to the flagging of Bifidobacterium.
The first document, from ANSES (2019), is the case report that led to the inclusion of Bifidobacterium on the list. It describes an episode of infectious endocarditis attributed to probiotic consumption in a 66-year-old woman in France. The second document (Esaiassen et al., 2017) is a series of 15 case reports of bacteremia caused by Bifidobacterium species. The third (Skiest et al., 1994) reports an individual case of anaerobic pericarditis in a 63-year-old woman in the United States. The remaining three documents (Brook, 2009; Brunser, 2017; Costa et al., 2018) are review papers citing 13 different articles that collectively describe 33 individual cases of infections and complications associated with various Bifidobacterium species across multiple regions of the world.
Taken together, these six documents, encompassing 50 individual cases, appear to have formed the basis for the suggestion that Bifidobacterium may be associated with infectious endocarditis (or bacterial infections).
Considering the cases and data returned
The problem with this conclusion arises when the publications are analyzed in more detail.
First, the apparent 49 cases attributed to Bifidobacterium show substantial overlap among the documents reviewed. Several reported cases are repeated across publications, and some are based on secondary references rather than original reports. In addition, some cases cited as evidence of Bifidobacterium-associated complications involve microorganisms other than Bifidobacterium, including cases where the individual was taking probiotics containing Bifidobacterium but the infection was attributed to other bacterial species.
Therefore, the initial impression of cases of complications associated with Bifidobacterium is reduced to 34 cases.
Is 34 cases a large number? This depends on the temporal, geographical and clinical context. From a temporal perspective, 34 cases do not represent a large number when considered over several decades, with reports spanning approximately 47 to 72 years. Geographically, the cases originate from multiple regions of the world, representing a large number of person-years of exposure.
Regarding clinical severity, deaths were reported in fewer than one-quarter of the cases. Importantly, the underlying health status of the individuals involved should also be considered. A substantial proportion of cases occurred in vulnerable populations, including pre-term newborns and adults with relevant risk factors such as recent delivery, bowel obstruction, metastatic cancer or severe underweight.
The clinical outcomes reported were primarily bacteremia (85%), followed by sepsis (9%), with septic arthritis and pericarditis each accounting for one case. It is also important to note that the methodology used in the report may group different clinical conditions, such as bacteremia, sepsis and pericarditis, under the broader category of infectious endocarditis/bacterial infection.
Another key question is whether the Bifidobacterium identified in these cases came from foods or supplements containing these bacteria. Probiotic use was not explicitly reported in 30% of cases, and molecular confirmation was lacking in nearly 70% of reports. Based on the available information, probiotic products containing Bifidobacterium may have been involved in only 40% to 50% of the 34 cases. In addition, in 20% of cases, patients were taking probiotics containing Bifidobacterium, but the bacteria recovered from infected tissues were not Bifidobacterium and were not considered the cause of infection.
In the 2019 ANSES case, for example, a patient was taking probiotic supplements containing six species of Bifidobacterium and Lactobacillus (eight strains in total), but only Lactobacillus species were recovered from tissues, with only one matching a species found in the supplements. Another reported infection involving Bifidobacterium was attributed to a medical procedure-related event rather than probiotic consumption.
What’s more, in nine of the 34 reports, where Bifidobacterium was detected in affected tissues, other microorganisms—including bacteria and fungi—were also identified, indicating that these were often complex, mixed infections rather than cases involving Bifidobacterium alone.
Avoiding disproportionate influence and confusion
When considering only unique, non-duplicated cases where (1) Bifidobacterium was unequivocally and exclusively responsible for an adverse event, and (2) food consumption is a plausible source of exposure (i.e., oral exposure meets the methodology’s inclusion criteria), only seven cases remain.
These were among the mildest events reported: no deaths occurred, with one case of sepsis, five cases of bacteremia and one case of septic arthritis in a patient taking a Bifidobacterium-containing supplement who had undergone hip arthroplasty, raising questions about the route of infection. No cases of endocarditis or pericarditis were identified. These seven cases are described in three publications included in the report (Costa, 2018; Brunser, 2017; Esaiassen, 2017). Applying the report’s methodology, Bifidobacterium’s score would decrease from 6 to 5.
Importantly, the case report that initially placed Bifidobacterium on the shortlist (ANSES, 2019) does not describe endocarditis caused by Bifidobacterium but rather by Lactobacillus species in a person consuming supplements containing both genera. Therefore, the basis for identifying Bifidobacterium as a potential infectious endocarditis risk is questionable.
The cited publications generally support caution in probiotic use, while recognizing the strong safety record of these products. However, ANSES (2019) states that no cases of endocarditis associated with probiotics containing Bifidobacterium spp. had been identified in the literature. Similarly, Brunser (2017) notes the rarity of Bifidobacterium isolation from blood cultures and that reported cases of Bifidobacterium-associated sepsis remain exceptional although the genus is extensively used in both the general population and in preterm babies at risk of necrotizing enterocolitis.
Overall, the inclusion of Bifidobacterium in the external scientific report should not, by itself, result in a change to its current regulatory status in European food supplements. The available evidence indicates that Bifidobacterium species and their use in foods have a favorable safety profile, while continued monitoring remains appropriate.
The shortlist should also be interpreted considering the methodology used to generate it, which is designed to identify potential risks with high sensitivity. Because the scoring system considers the number of documents rather than individual cases, isolated case reports may have a disproportionate influence on the ranking.
Finally, the terminology used has contributed to confusion, as infectious endocarditis appears to have been used interchangeably with bacterial infection. However, no documented case of Bifidobacterium-caused endocarditis supports its inclusion on this basis; the closest example is a polymicrobial pericarditis associated with an invasive medical procedure rather than food or supplement consumption.
