select search filters
briefings
roundups & rapid reactions
Fiona fox's blog

expert reaction to conference abstract looking at restricting dietary emulsifiers in people with Crohn’s disease

A conference abstract presented at the European Crohn’s and Colitis Organisation (ECCO) Conference looks at emulsifier restriction in people with Crohn’s disease. 

 

Dr Dominic Farsi, Postdoctoral Researcher in Nutrition Science, INRAE (l’Institut national de recherche pour l’agriculture, l’alimentation et l’environnement), France, said:

Does the press release accurately reflect the science?

“Yes.

 

Is there enough data and information available to be able to assess the quality of the science?

“Yes I believe there is enough information.  In addition, ECCO is one of the prominent organisations in gastroenterology and this abstract will have been reviewed by the conference committee (though that is different from journal peer review) and also the fact that it is being presented at the meeting would suggest that it is of a high standard as it has been accepted and chosen for the meeting.

 

How does this work fit with the existing evidence? + What are the implications in the real world?  Is there any overspeculation?

“A lot of the evidence to date surrounding emulsifiers and Crohn’s comes from rodent models, which while informative, will always have limitations, as, it is rodents, not humans.  This evidence does suggest that food additive emulsifiers may negatively impact the gastrointestinal environment, including the gut microbiota.  The evidence in humans is very limited and so this study could be viewed as foundational to help understand the potential impact of food additive emulsifiers in Crohn’s.  This study is a first of its kind and to date, we have not had significant data such as this, the results are positive and exciting, and as more research is performed in this topic, it will help further develop our understanding.  If future research confirms that making a dietary change such as removing emulsifiers from the diet can help people manage Crohn’s, then that would be viewed as very positive – but we need more evidence first.”

 

Prof Kim Barrett, Vice Dean for Research and Distinguished Professor of Physiology and Membrane Biology, UC Davis School of Medicine, said:

“This preliminary presentation reports that removing emulsifiers from the diet may be helpful in patients with Crohn’s disease.  It’s difficult to be fully confident in the conclusions since details of the study are not yet available, and more subjects dropped out of the control group vs. the treatment arm.  Nevertheless, the findings available are consistent with predictions from previous animal studies, and could offer a simple and inexpensive way to reduce symptoms in patients with limited treatment options.”

 

Prof Gary Frost, Chair in Nutrition and Dietetics, Imperial College London, said:

“I think that the study is well designed, it is clear how they have decreased exposure to emulsifiers to all in the trial then reintroduced them in a portfolio of food in the control group giving the low emulsifier diet (LED) group foods that had a low emulsifier content.

“The intention to treat results look clear as reported in the abstract but the protocol results suggests that about 60% of the control group did not finish the study there for the results are not so clear.  There needs to be more understanding about the reasons for the stark difference in not completing the trial.  (The results are expressed in two ways.  The intention to treat results include all the volunteers in the study even if they dropped out.  Although the authors do not say how they used the data usually the last recorded measure is taken as the final results.  This means if someone dropped out at 4 weeks the results at 4 weeks will be taken as the last results and analysed with others that have completed at 8 weeks.  It is this analysis that shows a significant difference.  The per protocol analysis only analyses people who completed the trial.  The problem they have is only 49 of the 75 controls completed the study.  There is no explanation for this level of patients not completing.)”

 

Prof Alexandra Johnstone, Theme Lead for Nutrition, Obesity and Disease, Rowett Institute, University of Aberdeen, said:

“There is currently a lot of interest in the role for ultra processed foods and health, and this study embraces this trend, to examine the role of emulsifiers, which are food additives, within a specific UK patient group.  Regulatory bodies ensure that food additives are rigorously tested for safety and additives continue to undergo long‐term monitoring for their effects on chronic health conditions.  Food additives that pass these safety tests are given an ‘E’ number which must be listed on packaging.  Whilst consumption of some food additives (e.g. artificial sweeteners) can be limited through food choice, it is much more difficult to avoid ingestion of emulsifiers (also known as surfactants or detergents) because they are commonly added to a wide variety of foods within the modern Western diets.  Some emulsifiers are man-made or synthetic (e.g. polysorbates) and some occur naturally (e.g. lecithin).  In this study, the emulsifiers that were provided, were carrageenan, carboxymethylcellulose and polysorbate-80 (control).  For example, carboxymethylcellulose is a non‐digestible polysaccharide polymer, hence its common use as a thickening agent and stabilizer in food emulsions.

“Emerging evidence1,2,3 suggests that permitted dietary emulsifiers may impact on gut health through impairing intestinal barrier function, thus increasing antigen exposure, and/or by modulating the microbiota, thus potentially increasing the incidence of inflammatory bowel disease (IBD) – Crohn’s disease is a form of IBD.

“The study to be presented at European Crohn’s and Colitis Organisation (ECCO) Conference in Berlin provides hope for patients with Crohn’s disease to manage symptoms.  The press release accurately reflects the brief data shared in the abstract, and I would welcome seeing the full dataset as a paper, once peer reviewed.  The limitations of the study are that this is a relatively small group of patients (N=154) and conducted over a short time-frame (8 weeks).  Patients with active flare in their disease were recruited, which was measured as inflammatory response, and the study abstract results indicate that the low emulsifier diet reduced inflammatory symptoms and promoted remission.  It was good to read the quote from the patient in the press release about the positive impact this had on quality of life.  Going forward, a peer-reviewed publication will likely reveal a more comprehensive analysis and discussion from the study team, including exploring the putative mechanism(s) of action.  For example, recognising the small subject numbers and limitations of the results that would need to be explored in a larger cohort of patients.  The study does offers hope for patients for future research on management of symptoms through dietary changes.”

 

1 Partridge D, Lloyd KA, Rhodes JM, Walker AW, Johnstone AM, Campbell BJ. Food additives: Assessing the impact of exposure to permitted emulsifiers on bowel and metabolic health – introducing the FADiets study. Nutr Bull. 2019 Dec;44(4):329-349. doi: 10.1111/nbu.12408.

2 Naimi, S., Viennois, E., Gewirtz, A.T. et al. Direct impact of commonly used dietary emulsifiers on human gut microbiota. Microbiome 9, 66 (2021). https://doi.org/10.1186/s40168-020-00996-6

3 Chassaing B, Koren O, Goodrich JK, Poole AC, Srinivasan S, Ley RE, Gewirtz AT. Dietary emulsifiers impact the mouse gut microbiota promoting colitis and metabolic syndrome. Nature. 2015 Mar 5;519(7541):92-6. doi: 10.1038/nature14232. Epub 2015 Feb 25. Erratum in: Nature. 2016 Aug 11;536(7615):238. doi: 10.1038/nature18000.

 

Dr Elizabeth Lund, Independent Nutrition Consultant, previously Research Leader at Quadram Institute of Biological Sciences, Norwich, said:

“This study looking at the potential benefit of diets low in emulsifiers shows a potential benefit for patients with mild to moderate Crohn’s disease.  It is only an abstract for a meeting and thus lacks detail and has not yet been peer reviewed.  This is important for journalists to recognise.  The nearly 20% drop out rate, that is the difference between those recruited and those completing the study, is on the high side but not atypical.  I would want to better understand why these people dropped out.  Nevertheless, the data looks very promising and – emphasising this data is yet to be peer reviewed – would suggest that this diet is well worth a try for Crohn’s patients, probably helping many if not all who suffer from this condition.  It is unlikely this is the only trigger for Crohn’s; inflammatory conditions tend to be very individual.”

 

Dr Baptiste Leurent, Associate Professor in Medical Statistics, UCL, said:

“This research is only published as a short summary, and it is difficult to assess its quality.  The results could be interesting, but their validity can only be assessed (and communicated to the public) once the full study will be published.

“I am particularly concerned by the data presented in Table 1, which suggest that the differences found could be related to assumptions made regarding missing data.

“The abstract does not state clearly how many participants had Crohn’s disease activity index (CDAI) data at 8 weeks, but the numbers reported suggest there may have been participants with missing CDAI response, particularly in the control group.  And that all of them may have been counted as “no CDAI response” in the primary (intention-to-treat) analysis.

“Given there was much more missing data in the control group, this would explain (at least some of) the lower CDAI response seen in the intervention group.  When looking at the `per-protocol’ analysis (participants with actual CDAI data) none of the results were statistically different between the two groups.

“Much other information is missing from the abstract to assess the study validity.  For example, it is not clear if the participants were aware of the emulsifier content of the food they were receiving, how similar were the two types of food received (apart from the emulsifier content), and to which extent participants ate the food they received.  Note that what is being evaluated here is receiving regular emulsifier-free food, compared to receiving regular high emulsifier content food.  It is not what is defined as LED (“low emulsifier diet”) in the abstract, which consisted of dietary advice and other material (both groups received this).  No protocol or analysis plan were published – it would be important to know if the outcomes reported here are all the outcomes that were originally planned to be looked at or if there were others.

“Further information is needed before these results can be communicated to the public, and I would be very concerned if any diet recommendation is made based on this abstract.”

 

Rachel Richardson, Methods Support Unit Manager, Evidence Production and Methods Directorate, Cochrane, said:

“It is difficult to properly assess this study as all we have is an abstract.  This means that the full report of the study is not available for scrutiny.  It would be helpful to have more detail in order to assess the validity and generalisability of these results.

“A few examples: the number of people who dropped out was high – only 73% completed the study, even though it only lasted for 8 weeks and the drop-out rate was higher in the control group than in the intervention group.  High rates of dropout can lead to bias in a study’s results, so it would be helpful to know more about why people dropped out.

“Another example is the fact that the authors adjusted the results for a variety of possible confounding factors, including BMI.  However this was a randomised trial which means that confounders should be evenly distributed between groups and no adjustments are necessary.  It would be useful to know why authors needed to do this.

“A final striking point is that we do not know from the abstract whether the results were precise enough to demonstrate meaningful improvement across all possible scenarios.  The confidence intervals, or margins of error, were quite wide for a number of the effect estimates.  Whilst improvement could be significant, there is also the possibility that the effect could be small, which may mean no clinical difference.”

 

Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University, said:

“These findings are interesting.  However, they are based on a conference abstract.  The full research report will not (yet) have been through the full peer review process, that it will need to go through before it can be published in a scientific journal.  Peer reviewers might (or might not) pick up important problems in the complete trial report.  Also, we don’t have by any means complete information, only a press release and a rather brief abstract (summary) of the findings.  So, while on the face of it the results look encouraging, it’s just not possible to evaluate the quality of the research properly on the basis of the information available.

“Overall, statistically, the results don’t seem to me to be nearly as clear-cut as the press release implies.  I’ll explain why I think this below.  It all rather depends on the reasons for the different rates of completing the full trial for the trial in the two groups (those who were on the diet without emulsifiers, and the control group who had emulsifiers).  The small amount of information we have doesn’t allow me to make a judgement on that.  Maybe the trial does show everything that the press release says it shows, but it remains possible that it doesn’t, not entirely anyway.  We need to know more.

“One issue is as follows.  As is usual in reports of randomised clinical trials, the researchers present two different sets of comparisons between the ‘intervention’ group (who had a diet without emulsifiers) and the ‘control’ group (who were given some emulsifiers).  One is the so-called ‘intention to treat’ comparison, which compared the results for everyone for whom results are available, regardless of whether they stayed on the diet that they were allocated to for the duration of the trial.  The other is the so-called ‘per-protocol’ analysis, that compares the results of only those patients who continued on the diet they were allocated to, for the full duration of the trial.  Usually the results of the intention to treat analysis are considered to be more revealing.  However, it’s important to consider the per-protocol results as well, particularly if a lot of people didn’t remain on the allocated treatment for the full period.  In this trial, about four out of five (81%) of those allocated to the emulsifier-free diet were on that diet throughout, and so contributed to the per-protocol analysis.  But a considerably smaller proportion, only about two out of three (65%), of those allocated to the control diet remained on it throughout.  The information we have doesn’t give an explanation for the differences in completion rates, and it might well matter.  Maybe people on the control diet were more likely to drop out because they felt it wasn’t helping them, or maybe they were eating more emulsifiers than usual and felt this was disagreeing with their digestion or health, or maybe something else – we don’t know.

“The results repeated in the press release are based on the intention to treat analysis.  But the results from the per-protocol analysis are less clear-cut, and indeed, apart from the finding about the proportion who said their symptoms improved, the other three outcomes given in the table in the Abstract don’t have large enough differences between the no-emulsifier and the control group to allow a conclusion that differences are definitely due to anything other than random chance.  (This doesn’t mean that they are due to no more than chance, just that it’s still feasible that they could be no more than chance.)

“Another issue is that the trial is not comparing what patients with Crohn’s disease do on a diet without emulsifiers, compared to what they would otherwise have eaten.  The control group of patients ate a diet containing emulsifiers, together with some snacks provided by the researchers, that also contained emulsifiers.  The patients in the intervention (no emulsifiers) group also were given snacks and some food provided by the researchers, but without emulsifiers.  I am no dietician, but I do know some people with Crohn’s disease, and they all avoid various types of food.  So the comparison in the trial seems to be between people definitely eating a special diet with no emulsifiers, with another group who are eating a different special diet that does contain emulsifiers.  This makes sense, to ensure, as far as practicable, that all the patients in each of the two groups are eating more or less the same things as the others in their group, but it does imply that the researchers are not comparing people on a special diet free of emulsifiers, with those eating what they would typically eat anyway.  You’d have to ask a dietician rather than a statistician to be sure of how important this is in the context of real-world diets.”

 

 

 

Abstract title: ‘Emulsifier restriction is an effective therapy for active Crohn’s disease: the ADDapt trial – a multi-centre, randomised, double-blind, placebo-controlled, re-supplementation trial in 154 patients’ by A. Bancil et al was presented at the European Crohn’s and Colitis Organisation (ECCO) Conference in Berlin. The embargo lifted at 17:20 UK time on Friday 21 February 2025.

There is no paper.

 

 

Declared interests

Dr Dominic Farsi: “During part of this study, I was within the department and involved in other diet trials, however not this one (the ADDapt trial).  I have since left this role and I am now a Postdoctoral Researcher at INRAE in France.  I have no other conflicts of interest.”

Prof Kim Barrett: “I have no conflicts of interest.”

Prof Gary Frost: “None.”

Prof Alexandra Johnstone: “AJ holds funding from UKRI which involves UK retail partners, the Scottish Government.  She currently holds voluntary committee position with Nutrition Society, Association for Nutrition and British Nutrition Foundation.”

Dr Elizabeth Lund: “No conflicts of interest to declare.”

Dr Baptiste Leurent: “No conflict of interest.”

Rachel Richardson: “I have no interests to declare.”

Prof Kevin McConway: “Previously a Trustee of the SMC and a member of its Advisory Committee.”

in this section

filter RoundUps by year

search by tag