Following an application from Danone Produits Frais France submitted pursuant to Article 14 of Regulation (EC) No 1924/2006 via the Competent Authority of France, the Panel on Dietetic Products, Nutrition and Allergies was asked to deliver an opinion on the scientific substantiation of a health claim related to a fermented milk drink containing Lactobacillus casei DN-114 001 plus yoghurt symbiosis (Actimel®) and reduction of the presence of Clostridium difficile (C. difficile) toxins in the gut which reduces the incidence of acute diarrhoea.
The scope of the application was proposed to fall under a health claim referring to disease risk reduction.
The food constituent that is the subject of the health claim is a fermented milk product (Actimel®) containing Lactobacillus casei DN-114001 and yoghurt cultures (Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus). The identification and characterisation of the strains have been performed by applying phenotypic and genotypic methods. The Panel considers that the food constituent, Actimel®, which is the subject of the health claim, is sufficiently characterised.
The claimed effect is “decreases the presence of C. difficile toxins (the risk factor), in the intestinal tract and reduces the incidence of acute diarrhoea associated with their presence in the gut of susceptible ageing people”. The target population is adults over 50 years old receiving antibiotic treatment. The Panel considers that reducing the risk of C. difficile diarrhoea by reducing the presence of C. difficile toxins is a beneficial physiological effect.
In total the applicant indicated seven publications on human studies, three unpublished human studies, eight published and one unpublished non-human studies to be pertinent for the claimed effect. In addition the applicant provided 65 references related to C. difficile, antibiotic associated diarrhoea (AAD), the homeostasis of the intestine and possible mechanisms of “probiotics”. These 65 references did not refer to Actimel® or its bacterial strains. Three human studies referred to C. difficile associated diarrhoea (CDAD).
In an open non-controlled observational study with 213 elderly patients with a mean age of 88 years in two geriatric care wards of a hospital, subjects received twice daily Actimel® for the duration of an antibiotic treatment course and 7 days after. The study recorded the incidence of CDAD among patients in the geriatric wards for the period November 2007 to January 2008 and compared it with the historical incidence in the period of November 2006 to January 2007 when this yoghurt drink had not been consumed. The Panel notes that this study was not controlled for factors other than the consumption of Actimel® that might have influenced the incidence of CDAD. The Panel considers that no conclusion can be drawn from this study for the scientific substantiation of the claimed effect.
A randomised, double-blind, placebo-controlled pilot (RCT) study was designed to test the feasibility of a daily intervention with 2 x 100 ml per day of Actimel® and of a non-fermented acidified dairy drink to investigate the effect on the occurrence of AAD and of CDAD in elderly hospitalised patients receiving antibiotics (10 per group). The applicant indicated in the study report that “no statistical comparison was performed in the context of a small pilot study”. The Panel agrees that this pilot study was underpowered to assess the effects of Actimel® on the incidence of AAD and CDAD. The Panel considers that no conclusions can be drawn from this study for the scientific substantiation of the claimed effect.
A RCT with 135 patients of 50 or more years of age under antibiotic treatment studied the effect of daily 2 x 100 ml of the yoghurt drink Actimel® compared to 2 x 100 ml daily of a control product on the occurrence of AAD as the primary endpoint and CDAD as a secondary endpoint. Test and control product were to be consumed twice a day during the course of antibiotic treatment and for one week thereafter. As the products were packaged in bottles of different size (100 g Actimel® bottle versus 200 ml Yazoo® bottle), the hospital pharmacies removed the commercial labels and nursing staff were instructed to pour 100 ml of the respective products into a cup and to deliver to the patients in order to ensure blinding. According to the study protocol, patients who were discharged from the hospital before the end of the intervention were given adequate supplies of test and control products with instructions for use labelled on the bottles. The Panel considers that while the blinding procedure used in the hospital was adequate, the procedure used following discharge could have resulted in un-blinding of patients due to the different size of the bottles for the two products.
“Diarrhoea” was defined as more than two liquid stools a day for three or more days in quantities in excess of normal for each patient. According to the study protocol, stool samples of all patients should have been tested at baseline, when diarrhoea was reported and at the end of the trial. The follow-up of 12 subjects of the Actimel® group and of 10 subjects of the control group could not be completed. In a complete- case analysis (per-protocol analysis), the authors reported AAD for 7 out of 57 patients in the Actimel® group and for 19 out of 56 patients of the control group. In none of the patients in the Actimel® group and in 9 out of 53 of the control group diarrhoea was associated with a stool sample positive for C. difficile toxins. This corresponds to a statistically significant 22 % (95 % CI: 7-37) and 17 % (95 % CI: 7-27) absolute risk reduction for the occurrence of AAD and CDAD, respectively. The Panel notes that C. difficile toxins were only measured in patients who developed AAD and that there were no data on the effect of Actimel® on the occurrence of C. difficile toxins in the faeces of patients without AAD. On request by EFSA, the applicant provided a re-assessment of the data with and without data imputation for the non-completers. The applicant provided a re-assessment with data imputation assuming that a) none of the 22 drop outs had AAD and b) that 20 % in both groups had AAD. The results remained statistically significant (p < 0.05) in favour of the test product for both scenarios. The Panel notes that statistical significance is sensitive to the treatment of the missing data and that no data imputation has been applied for more conservative scenarios than 20 % assumed occurrence of AAD.
The Panel notes a number of weaknesses in this study: potential bias through un-blinding of the patients to the products; the lack of data on the presence of C. difficile toxins (the risk factor) in the faeces of patients at the end of the intervention; no scenario has been applied for the missing data imputation for more conservative scenarios than 20 % assumed occurrence of AAD. The Panel considers that the weaknesses of this study limit its value as a source of data to substantiate the claim.
An in vitro study was provided on the capability of Lc DN-114 001 to inhibit the growth of C. difficile. According to the summary report the supernatant of Lc DN-114 001 was tested with the “overlay method” modified by Danone, for its capacity to inhibit the in vitro growth of Salmonella thyphimurium, Listeria monocytogenes and C. difficile. According to the data provided, the supernatant of Lc DN-114 001 tested positive for its capability to inhibit the growth of all three pathogens. The Panel considers that the results of this in vitro study do not predict the occurrence of such an inhibitory effect in humans.
Five human studies investigated the kinetics, metabolism and survival of Lc DN-114 001 and its impact on the human microbiota during its gastrointestinal transit. These studies demonstrated that Lc DN-114 001 can partially survive the gastrointestinal transit. A study demonstrated the expression of 10 genes of Lc DN-114 001 in human faeces after gastrointestinal transit and did not find evidence for multiplication. Results from animal studies showed that Lc DN-114 001 can partially survive gastrointestinal transit in human flora-associated mice and rats and is excreted without detectable multiplication with the same kinetics as the inert transit marker Bacillus subtilis spores. The Panel notes that human and animal data consistently showed partial survival of Lc DN-114 001 during its gastrointestinal passage without detectable multiplication.
Regarding possible mechanisms by which Actimel® could reduce the risk of C. difficile diarrhoea by reducing the presence of C. difficile toxins, the applicant suggested that Actimel® increased colonisation resistance against pathogens and inhibited pathogen proliferation by modulation of the intestinal microbiota, reduction of the gut epithelial permeability, modulation of gut epithelial receptors and induction of specific IgA production.
The applicant provided two human studies on the effect of Actimel® on gastrointestinal infections and gastrointestinal pathogens, but which did not refer to C. difficile. In a RCT with 1072 elderly volunteers consuming 2 x 100 ml Actimel® per day over 3 months there was no difference in the incidence of common infectious diseases, the incidence and duration of gastrointestinal infections, the occurrence of fever, the occurrence of pathogens, the number of prescribed medication and immunological parameters. In another RCT, 250 young adult volunteers enrolled in firefighting training consumed 2 x 100 ml Actimel® per day over 7 weeks. There was no difference in the incidence of common infectious diseases, the incidence of gastrointestinal infections, the occurrence of fever, the occurrence of pathogens in faeces and the number of prescribed medication. The Panel notes that no information has been provided on C. difficile. The Panel considers that these two RCTs do not support an effect of Actimel® on the reduction of gastrointestinal infections or gastrointestinal pathogens in humans.
Three other human studies addressed effects on the intestinal microbiota. In an uncontrolled human trial with 12 healthy subjects no statistically significant change occurred after 10 days of daily consumption of 300 ml Actimel® in either the dominant members of the faecal microbiota, bacterial enzyme activities, pH and metabolites such as short-chain fatty acids. This was confirmed by another human trial in 7 volunteers after 8 days of daily consumption of 300 ml Actimel® which did not find any statistically significant change of the proportions of 7 phylogenetic groups . Also a study on the effect of a daily intake of 125 g Actimel® in 39 infants and young children of 10 -18 months of age, did not find any statistically significant modification of the number of total anaerobes, bifidobacteria, bacteroides, and enterobacteria, bacterial metabolites and bacterial enzyme activities in faeces at the end of a 30-day intervention. Based on the parameters measured, the Panel considers that the three human studies do not provide evidence that Actimel® can modulate the human gut microbiota.
The applicant provided another two animal studies and three in vitro studies on possible effects of Actimel® or Lc DN-114 001 in model systems related to immune function and infection. The Panel notes that none of these five studies investigated C. difficile. The Panel considers that the evidence provided does not establish that effects of Actimel® or Lc DN-114 001 observed in these animal and in vitro studies can predict the occurrence of effects on immune function or infection in humans.
In weighing the evidence, the Panel took into account that human and animal studies showed partial survival of Lc DN-114 001 during its gastrointestinal passage, that one human intervention study with Actimel® which showed a statistically significant risk reduction for CDAD had considerable limitations, that there were only limited data on the effect of Actimel® on the reduction C. difficile toxins (the risk factor) in humans, that one study which showed an inhibitory effect of Lc DN-114 001 on the growth of C. difficile in vitro does not predict the occurrence of an effect against C. difficile in humans, that five further human studies do not support the proposed mechanisms by which Actimel® could exert the claimed effect, and that the evidence provided from a further two animal and three in vitro studies does not establish that effects of Actimel® or Lc DN-114 001 in these model systems related to immune function and infection can predict the occurrence of such effects in humans.
The Panel concludes that the evidence provided is insufficient to establish a cause and effect relationship between the consumption of Actimel® and reduction of the risk of C. difficile diarrhoea by reducing the presence of C. difficile toxins.