select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
A study published in the Lancet Diabetes & Endocrinology looks at food additive emulsifiers and risk of type 2 diabetes.
Prof Pete Wilde, Emeritus Fellow, Quadram Institute, said:
“It’s an interesting paper where they are able to segregate different additives and their impact on type 2 diabetes, with intakes of some giving relatively strong correlations and others show little impact. However, I think the authors have missed an opportunity here. Most of the additives studied are not actually emulsifiers. There are a range of additives listed, some are emulsifiers, some are thickeners/stabilisers and some are acidity regulators.
“Emulsifiers such as the various mono- and di-glycerides, lecithins, polyglycerol esters are small amphiphilic molecules which are composed of a hydrophilic part and a hydrophobic part. They sit at the interface between oil and water, reducing the interfacial tension between the oil and water and help to emulsify the oil into the water.
“Many of the molecules studied in this paper are actually thickeners. Alginates, carageenans, cellulose, starches and gums. They are all large polymeric molecules which absorb water and used to increase viscosity and thicken foods. They are often called stabilisers as the viscosity stops oil droplets and other particles from moving around and separating out. They act in a completely different way to true emulsifiers, therefore their mechanism of action in terms of health impact is likely to be different. The only thickener which is sometimes used as an emulsifier is gum arabic.
“Some of the other additives studied include sodium bicarbonate which makes food less acidic, sodium citrate which makes food more acidic and various other salts. None of these act as emulsifiers or thickeners.
“If the authors had classified the additives according to function and molecular structure, they have been able to gain deeper insights into the possible mechanisms involved in order to develop more mechanistic research hypotheses.
“Having said that, within additive classifications there is also significant range of effects, for example, of the true emulsifiers, only the acetyl tartaric acid esters of mono and diglycerides showed a significant increase in risk, and this is widely used in breadmaking. All of the other molecules listed which showed a significant relationship are in fact thickeners or acidity regulators. The exception being gum arabic, which is a thickener, but also has some emulsifying properties.”
Dr Sarah Berry, Reader, Department of Nutritional Sciences, King’s College London, said:
“This type of large-scale epidemiological study is a vital part of the scientific process. However, these studies cannot prove that emulsifiers cause type 2 diabetes. Because products that contain emulsifiers often contain a multitude of other ingredients, disentangling the effects of each compound is challenging.
“However, while emulsifiers are individually carefully checked for safety, few studies investigate how cocktails of these chemicals impact our health. So, this type of study should inform future, more tightly controlled clinical trials.
“With that said, existing evidence shows that a diet rich in the types of foods that contain emulsifiers, which have typically been heavily processed, is linked to poorer health outcomes. So, cutting down on these foods is likely a healthy choice.”
Dr Duane Mellor, Registered Dietitian and Senior Lecturer, Aston Medical School, Aston University, said:
“This is an interesting further analysis of the NutriNet Sante cohort which has reported on the association between food additives and ultra-processed foods and risk of health conditions including diabetes before. So, perhaps it is not surprising that this study and group of researchers find that emulsifiers which are common additives in manufactured foods are associated with risk of diabetes in a population that has previously shown a link between highly processed foods and diabetes.
“What needs to be considered is that the range of emulsifiers included in this study are quite varied, and includes sodium bicarbonate as an emulsifier, which is one of its uses, but it is perhaps more commonly used as a raising agent in the form of baking powder – so it seems that there are some overlap between how ingredients can be classified as they can have different uses in cooking as well as commercial food manufacture.
“The way the study estimated intake of emulsifiers needs to be treated with caution, as the study looked at three 24 hour food recalls and then looked to work back to estimate intakes of emulsifier ingredients which are used in minute quantities in foods. This is likely to mean the level of precision suggested in the paper is simply not possible and the paper does not clearly highlight any potential margin of error.
“What this paper does not fully consider is the difference in how the human body might process and manage emulsifiers – lecithins are a common emulsifier (which was not linked to risk of type 2 diabetes in this study) and is found naturally in egg yolks and is used to make mayonnaise from oil and water, which when mixed is an emulsion! Lecithins contain a range of compounds including choline, which in our body can be used to make acetylcholine which nerves use to transmit messages through our nervous system and for some people their bodies struggle to make enough of it and may require dietary sources of choline. This is very different to polysorbates which have no function in our bodies and in mechanistic studies have been linked to metabolic changes in rodents. So, it is important not to consider all emulsifiers as being the same.
“When the data is looked at to see the major sources of the emulsifiers associated with greatest chance of increasing the risk of type 2 diabetes, which was phosphates, the main source of these were cakes and biscuits – so it is perhaps more plausible that confounding factors are present in the data, as the researchers only sought to control for individual dietary components and ultra processed food intake; they did not consider overall dietary pattern quality and how healthy it is using a healthy dietary pattern score.”
Prof Gunter Kuhnle, Professor of Nutrition and Food Science, University of Reading, said:
“This is one of several studies from the NutriNet-Santé cohort that investigates the link between food additives and ill-health. Like all observational studies in nutrition research, it suffers from the difficulty of estimating food intake reliably. Foods are not standardised and manufacturers are not required to provide detailed information on additive use – it is therefore impossible to estimate the amount of emulsifier consumed and the results of the study are therefore very unreliable. If anything, the study can show an association between diabetes and foods that often contain certain emulsifiers, but it is unable to show associations with emulsifiers themselves.
“The authors investigate a large number of emulsifiers, which makes a chance finding very likely. The observed effect sizes are very small and unlikely to be meaningful. The study identifies phosphates as one of the emulsifiers that is associated with increased disease risk – but phosphates are naturally present in many foods, especially dairy and meat and, and natural phosphates are often the main source of intake. In this study, it appears that the main source of added phosphates are cakes and it is not difficult to assume that it may be rather intake of high fat high sugar foods (like cakes) that are associated with an increased diabetes risk and not the additives.
“In summary, the study is, by design, unable to provide reliable information about the effect of emulsifiers on health, and it does not claim to do so. Notably, other studies have shown that foods that contain emulsifiers – such as yoghurts – are associated with a reduced risk of type 2 diabetes.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University, said:
“This is, I believe, the third of a series of studies on associations between noncommunicable diseases and consumption of emulsifiers in food and drink, based on data from the major French long-term NutriNet-Santé cohort study. (The previous two looked at associations with cardiovascular disease and with certain cancers, and appear in references numbered 15 and 16 in this new paper. They were published in 2023 and earlier in 2024.)
“This new study has similar strengths and issues to the previous two. In statistical terms, it’s a well-conducted study of its type. The researchers found that participants who ate higher than average amounts of certain emulsifiers had a higher chance of developing type 2 diabetes.
“This did not apply to all the different emulsifiers they considered, by any means. They looked at over 60 emulsifiers, but found associations with diabetes risk only for seven of the individual emulsifiers and for one group of emulsifiers (carragenans). For some emulsifiers, this was because they had plenty of data from the participants but no substantial association between consumption and diabetes risk was found, though for others it was essentially because consumption of the emulsifier, amongst the participants, was too low to produce meaningful results. However, they are certainly not saying that all emulsifiers are associated with diabetes risk.
“But the study is observational. As the researchers acknowledge, this means that it can’t on its own establish whether or not the higher diabetes risk is caused by the higher consumption of the food additives concerned. They write, “A single observational epidemiological study is not sufficient per se to establish causality.” And they point out that this is the first study that examined emulsifier consumption in this much detail, so there aren’t yet other comparable studies to compare it with.
“The issue about cause arises because participants who consumed different amounts of the emulsifiers inevitably differed in lots of other ways too, and the differences in diabetes risk might be caused by these differences in other factors rather than by the consumption of the emulsifiers in question. The researchers made statistical adjustments to try to reduce the impact of several of these other factors, but in this kind of study, it’s impossible to be sure that everything important has been taken fully into account.
“The researchers also point out other limitations of their findings. They mention possible issues in measuring emulsifier intake, and problems in disentangling any possible effects of emulsifier additives from possible effects of other types of food additive. They point out that the participants in the NutriNet-Santé study are not very typical of the French population – four out of five are female, and overall the participants are more health-conscious than average (otherwise they may not have volunteered to take part in the study). So it’s unclear how far these findings generalise to wider populations of people, even to all French adults, let alone to people from other societies with different eating habits.
“Any study of this general type does not provide direct information on how food additives might cause illness, if in fact they do cause illness, in terms of what actually happens in the body. To investigate that would require more research, and indeed research of a different nature, involving perhaps studies in animals or in cell cultures. I’d say that’s always important, alongside large-scale observational studies like this one. But I think it’s particularly important in a context like this, where the researchers looked at a wide range of emulsifiers that differ in their chemical composition and structure, so that it’s possible that they work in the body in different ways. The researchers do recommend that further research should be done.
“The study certainly has strengths too. For instance, it had a large number of participants (over 100,000) and followed them up for almost 7 years on average. The measurement of consumption of the food additives in question was careful and detailed.
“In case you wonder why researchers would use this type of large observational study, given the issues about cause and effect, measurement, and lack of findings about exactly what is going on in terms of internal mechanisms, the reason is that they add information that can’t easily be found in any other way. Because noncommunicable diseases like type 2 diabetes (or cancers or cardiovascular disease) take a long time to develop, a study needs to run for years in order to pick up enough cases of the disease to get good risk estimates. But, for example, a randomised trial where the researchers determined how much emulsifier people eat, rather than letting them choose for themselves what they eat (as in this new study), would need to run for a similar length of time to the 7-year average length of participation in this study, making it infeasible and probably unethical. In order to tie things down properly in terms of cause and effect, one needs more than just this one large, detailed, cohort study, and also a lot of evidence from other types of research.”
‘Food additive emulsifiers and the risk of type 2 diabetes: analysis of data from the NutriNet-Santé prospective cohort study’ by Clara Salame et al. was published in the Lancet Diabetes & Endocrinology at 23:30 UK time on Tuesday 23 April 2024.
DOI: 10.1016/S2213-8587(24)00086-X
Declared interests
Prof Pete Wilde: “I have received no personal funding from the food industry in the past. I do not have any current funding from industry. The vast majority of my funding has been government (UKRI) or EU. In the past I have received partial support for my research from a number of food companies in the past, including Unilever, Mondelez, Nestle and Quorn.”
Dr Sarah Berry: “I receive options and consultancy from ZOE Ltd.”
Dr Duane Mellor: “Duane has discussed the topic of ultra-processed foods and food reformulation with the APPG Obesity, EUFIC and several food companies.”
Prof Gunter Kuhnle: “https://kuhnle.co.uk/coi.html
Employment
Director of the University of Reading Chemical Analysis Facility with frequent interaction with instrument manufacturers and external and internal users from a range of backgrounds.
Research funding
Committee membership
Committee on Toxicity (COT) and various working groups
past member of the EFSA ANS panel and the American Academy of Nutrition and Dietetics working group on flavanols; Scientific advisory committee of the British Nutrition Foundation; Organising committee of the International Conference on Polyphenols and Health.
Scientific organisations
Others
Trustee of a Parent Teacher Association.
Family Vineyard.”
Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee. My quote above is in my capacity as an independent professional statistician.”