Interview Summary

Resources for readers:

• Rios-Leyvraz, Magali; Montez, J. “Health Effects of the Use of Non-Sugar Sweeteners: Systematic Review and Meta-analysis,.”
  
• WHO guideline on use of non-sugar sweeteners
  

This is really important work, so let’s get right down to it. So, Francesco, let’s begin with you. Would you help our listeners understand why the WHO, developed this guideline, and what do you hope to accomplish now?

Francesco: So our interest in the topic really came after the release of the WHO Guideline: Sugars Intake for Adults and Children in 2015. That recommendation was, of course, to reduce the amount of sugar to less than 10% of energy, and better if less than 5% for weight control and for dental hygiene. At the time, there was an interest, particularly from the food and beverage manufacturers, to consider the use of non-sugar sweeteners as part of the interventions in reducing sugar intake. This has been taken up by consumers since, and the availability of non-sugar sweeteners has increased. We thought it was important to review the evidence in a systematic manner, and to see whether this move was actually achieving the intended aim to control weight gain. So we thought of looking at non-sugar sweeteners use through the WHO guideline development process. Our hope, first of all, is to make sure that people achieve their intended health benefit, but also to avoid an unintended consequence on health.

I remember how much impact that report on sugar had, back when WHO released it. And boy, the same thing seems to be happening with the guideline that We are talking about today. So Jason, can you walk us through the key findings of the systematic review?

Jason: Sure. So the top line results of the systematic review are that in short term randomized controlled trials mostly lasting three months or less, those consuming higher amounts of non-sugar sweetener, compared to lower amounts or no non-sugar sweetener, has slightly lower body weight. So around 700 grams. And, slightly lower BMI: about 0.2 units. However, in long-term prospective cohort studies, many with several years of follow-up, body weight was either unchanged, or greater in those consuming higher amounts of non-sugar sweeteners. In addition, associations were also observed between non-sugar sweetener intake and increased risk of type 2 diabetes, cardiovascular diseases, and premature mortality. For diabetes, for example, we found about a 25% increase in risk. In RCTs – markers of these diseases – such as blood glucose, insulin, cholesterol, blood pressure, were largely unaffected by non-sugar sweetener use. In addition, we saw an increased risk of bladder cancer coming from case control studies and it was primarily driven by saccharin. Also, a 25% increase in risk of preterm birth in pregnant women in terms of body weight, and the risk of non-communicable diet-related diseases. We essentially have seemingly opposing results between the short-term randomized controlled trials and the long-term cohort studies.

Those are really concerning findings. Somebody in their own calculus might say, “Well, I’m willing to accept the risk for those things as long as it helps with my body weight.” But, you are talking about greater weight in people using the non-sugar sweeteners.

Jason: Right. In terms of long-term impact on body weight, there is not a lot of evidence. This is all coming from cohort studies, and cohort studies have caveats associated with them. But, what we see in those studies is it is either really not much of an positive impact, but we did see some increased risk. For example, there was an increased risk of incident obesity in one study. They looked at people at baseline without obesity using standard WHO cutoffs. Then they followed them up over time. They counted the number of people who transitioned over the obesity threshold, and they found that it was higher in those using non-sugar sweeteners. So there is some evidence, but I think we need additional evidence to be really sure about that in the other associations.

So Jason, you alluded to this earlier that there were differences in the health outcomes reported in randomized controlled trials versus observational studies. Can you explain why this might be?

Jason: Yes. So it is a bit nuanced and I think a lot of it has to do with the nature of the studies. To really understand this, we need to unpack that high-level evidence a bit. I think I’ll start just by talking about what is maybe the elephant in the room and that is reverse causation because it has been frequently invoked as a reason for the associations observed between non-sugar sweeteners and body weight and disease outcomes. Basically, reverse causation suggests that those already at elevated risk of disease initiated or increased the use of non-sugar sweeteners because of their risk status, rather than non-sugar sweeteners leading to the increased risk in otherwise healthier low risk individuals. That would make sense in this case. However, researchers who conducted the individual studies that we included in the review were also well aware of the potential for both reverse causation and residual confounding. Most of them really went to great lengths to minimize the possible interference of these phenomenon and they adjusted extensively for confounding variables and all the lifestyle variables. Quite often, they adjusted for other dietary variables. They looked at body weight, they stratified, and they did sensitivity analysis where they would limit the analysis to individuals with normal body weight. They also removed from the analyses, in some cases, those at risk for disease at baseline. Many of them did a sensitivity analysis where they excluded the first several years of assessing the outcome, just to make sure that someone who wasn’t already really at risk to develop a disease right away was not part of the analysis. You know, in some cases the effect was attenuated. But, for the most part, the associations were not affected. Certainly, that is the case for type 2 diabetes. So I think that reverse causation might have contributed, but, really the lengths that the study authors took to address it I calms us a bit in terms of whether or not We are worried about it.

It is very helpful to know that. You know, we’ve gone for years and years thinking that these things are safe, or at the worst have neutral effects. Now the studies are showing the negative effects. Partly, it is just because the science has marched ahead. There are many more studies now. People have been more careful with the design of the studies. More sweeteners have been studied. This seems like just kind of a natural progression of the scientific process.

Jason:  Yes, that is how it works. We continually add. There is still a lot to look at in terms of the differences. Another really important thing to remember is that in the RCTs, a variety of interventions were employed. A lot of the studies took people who were consuming a normal diet, and they just gave them, for example, a sugar-sweetened beverage, or non-sugar sweetened beverage. They wanted to look at what the effects of a sugar-sweetened beverage are. Very few actually looked at individuals who were habituated to sugar, mostly sugar-sweetened beverages. They then asked study participants to replace the sugar with non-sugar sweetener. When you look at just those small number of studies, the results really get attenuated for both body weight and for BMI. That is another reason why if you look at it that way, the results between the long-term and short-term tend to gravitate towards each other.

One last thing – I don’t want to overdo it here, but it is important – there are likely big, big differences in terms of how non-sugar sweeteners are used in these highly experimental randomized controlled trial settings and cohort studies, which are really more like real world use. In a randomized controlled trial, you’re instructed to do this, or that. Drink this, take that. You get quite a bit of support from the study team. You’ll get counseling, follow up, etc. Everyone knows they are in a scientific study. Quite often, they actually know which arm they are in. Whereas in the cohort study, in the real world, people use non-sugar sweeteners in a variety of ways. Sometimes they do it as a conscious replacement for sugars. But quite often, they just consume beverages containing non-sugar sweeteners or foods containing non-sugar sweeteners because they have a diet label on them, and they just perceive that those types of products to be healthy. Sometimes they use them as a way of having something sweeter or something that maybe they shouldn’t have later on in the day. It is much more complex in the real world and obviously the real world is how these things are used. So how sweeteners are used also can contribute to these differences between what we see in highly experimental randomized controlled trials and the real world.

So how do these non-sugar sweeteners cause these negative health effects?

Jason: We don’t know for sure. Let me just say that upfront. But we have ideas and the scientific community has ideas. A lot of mechanisms have been put forth to explain how this might be happening. A lot of them are actually linked to the expression of sweet taste receptors. They are expressed in the mouth, obviously, that is how we perceive things to be sweet. But they are also expressed elsewhere in glucose sensing cells, such as the gastrointestinal tract and pancreas. So there are a number of different ways that this might work. There could be effects on taste perception, sweet taste preference, a threshold of sweet taste sensitivity, and eating behavior as it impacts on hunger or appetite. There could be other neural responses. There is this whole discussion about the hedonic response to any sweet taste. Also, there might be release of metabolic hormones, and other biological molecules, in response to non-sugar sweetener intake. Then, of course, There is a lot of buzz around potential alterations to the gut microbiome, which is a really interesting area. But There is a lot of understanding still to be gained there. Also, as noted, in terms of potential mechanism, it could be that the behavioral component we just spoke about, in terms of how people actually use non-sugar sweeteners in the real world, could also contribute to the association between non-sugar sweetener use, and body weight, and disease outcomes.

A vast number of effects happen once these things get into the body. So Francesco, let me come back to you. What is WHO recommending? I know some recommendations that have been made are conditional. What does this mean?

Francesco: So the recommendation reads that the WHO suggests that non-sugar sweeteners not be used as a means of achieving weight control, or reducing the risk of non-communicable diseases. It is a very simple one. It is basically saying that you’re using the sweeteners, but they are not going to help you in weight control, or reducing the risk of that related non-communicable disease. So Kelly, as you said earlier, would I take the risk potentially associated with non-sugar sweeteners if I knew that this would benefit my weight? But in this case, what we are saying is, “Well, look, it is not going to benefit your weight control.” We are leaving it there, because the purpose of this review was, particularly, to look at weight control. We were not looking at other aspects of safety – which is actually the task of other regulatory/advisory bodies, although we had to consider in the recommendation when we called the balance of benefits and harm. So here, the recommendation was clearly to say, “Well, you shouldn’t use it, because there is no demonstrated benefit, and there might be some harm.”

Now, WHO can make two types of recommendations. Recommendations can be strong or conditional. The decision on whether to go for the stronger condition is based on a number of factors – certainty in the supporting scientific evidence, and the balance of desirable and undesirable consequences. A strong recommendation is the recommendation for which WHO gathering group is confident that the desirable consequences of implementing the recommendation outweighs the undesirable consequences. In most situations, this would be something that government can take on immediately and use in their policymaking. A conditional recommendation is a situation where the WHO guide group is less certain about that balance between the desirable and undesirable consequences. Therefore, we are basically saying, you need to consider the context, and then decide what to do with this particular recommendation. So, in this case, why did we go for conditional? There is an element related to the balance and the evidence that Jason described earlier on. There is also an issue about short term versus longer-term consumption. We have been trying to explain why not using does not create a problem. The action also depends on the context of use of this particular product. For example, the levels of consumption in different age groups in the population. So, in a sense, our understanding of the issue, is still not complete. We believe that we need to have more science on the topic. We decided to take a bit of a conservative approach. That is why we end up saying conditional, which means we still recommend that people not use non-sugar sweeteners. But in terms of making use with this recommendation, we would like governments to also do their own assessment on how to translate this into policy practice.

So Francesco, on one hand, we have sugar being a bad actor and now the replacements for sugars can be bad actors. So how does this get balanced out? Should people go back to sugar? Should they just get used to things being less sweet? How does this work?

Francesco: So definitely people should not go back to sugars because we know that consumption of free sugars has been linked to unhealthy weight gain, type 2 diabetes, cardiovascular disease, and dental caries. We definitely maintain the recommendation that we should limit free sugar intake. But unfortunately, non-sugar sweeteners are not producing a good benefit if we replace them with sugar. What they are saying is that both free sugars and non-sugar sweeteners should be limited. So yes indeed we would need to adjust our own taste to consuming products which are less sweet. We should also ask the producers to reduce the sweetness in their product.

So Jason, do you have some thoughts you’d like to share about this?

Jason: Yes. Again, from a scientific evidence standpoint, it is interesting. If you look at the literature, and you look at the studies, and we have lots of studies now, and a lot of systematic reviews, and umbrella reviews that really tried to make sense of all of this. But if you look at some of the impacts, associations between sugar intake, and diet-related noncommunicable diseases, especially type 2 diabetes, and you look at the actual magnitude of associations, the increased risk, they are really in line with what they are for non-sugar sweeteners. In fact, some recent modeling published in 2022, modeling in the prospective cohort studies, shows that when you replace sugar with non-sugar sweetener, you basically end up with a no result. Which means that non-sugar sweeteners, at least in terms of type 2 diabetes, are just as bad for you as sugar-sweetened beverages. These studies are all looking at beverages. It is interesting that we have to keep in mind that they may both be bad actors as you indicated, Kelly.

What an amazing picture. Francesco, so do you think that non-sugar sweeteners should not be used as a harm reduction mechanism to help people reduce their sugar intake?

Francesco: Yes, that is what our recommendation basically says. We are encouraging individuals, but also government, to think of other ways to reduce sugar intake that does not require the use of sweeteners as a replacement. There are ways to do that. I mean, certainly, there are unsweetened variants of the same food or drink. There are alternatives in diet patterns. You don’t necessarily have to have a dessert at every meal. You can have fruit. We know that the fruit sugars do not have the same metabolic effect as free sugars. We know that, for example, certain countries, like in the UK have given options to companies to use sweeteners instead, to avoid the sugar levy. Probably, we would ask those countries to have an assessment of what has happened and see whether that policy is still a useful one, or if it should be revised.

So Jason, I’d like to ask you about diabetes. Does the recommendations apply to people with diabetes, and can people with diabetes use these non-sugar sweeteners?

Jason: I’ll tell you, that is a question we are getting a lot of. Our guidance really is for preventing the development of disease, and not for treating, or managing existing disease, including diabetes. In terms of the process, making recommendations for preventing disease is very different than making recommendations for managing disease. Different scientific evidence needs to be reviewed. In the case of diabetes, some of the evidence is overlapping. But we really do need to look at a different body of evidence and a different assessment of potential risks and benefits needs to be made. This is all in the context of a larger plan of treatment, which may also include medication, and other lifestyle changes. Therefore, managing disease requires greater input from clinicians. Something else to remember is that individuals without diabetes do have the option to include some sugar in the diet when something sweet is desired. But for many with diabetes, this isn’t really an option. With that said, individuals with diabetes can certainly manage their disease without the use of non-sugar sweeteners. However, such decisions really should be made on an individual basis, and only in consultation with one’s healthcare provider.

So Jason, let me ask you a question about Stevia in particular. So there are some thinking that because it is natural, unlike the artificial sweeteners, it is somehow different. So why doesn’t the recommendation make distinctions between different non-sugar sweeteners?

Jason: That is a great question. Although all non-sugar sweeteners have unique chemical structures, and consequently may differ in their sweetness intensity, organoleptic properties, and routes of processing by the body, we don’t make a distinction in our guidance between individual sweeteners or between so-called natural sweeteners, like Stevia, and artificial sweeteners. This is because all sweeteners, including sugar, use the same biological pathway when it comes to how we perceive them as sweet tasting, so the sweet taste receptor I mentioned earlier. And, consequently, how they might impact weight and risk of disease. Also relevant is the possibility that we discussed before, which is that people may use these in ways that promote weight gain and disease, irrespective of what type of non-sugar sweeteners being used. We feel that the evidence for a class effect is sufficient to include them together in our recommendation.

That is very helpful to know. So Francesco, some organizations, for example, the Food and Drug Administration in the US, have declared over the years that these non-sugar sweeteners are safe for use. If those organizations feel they are safe, then why is WHO recommending that they not be?

Francesco: In a sense, it is a different issue. Before any of these compounds are used in food, it has to go through this safety assessment. You mentioned the Food and Drug Administration. FAO and WHO are managing a joint expert committee on food additives that is providing advice to the Codex Alimentarius. And all these bodies, basically what they are saying is, they look at the toxicology of the products and they use animal studies. In some cases, they look at human studies. But they are basically looking at acceptable daily intakes that provide usually short term impacts. We are also encouraging these bodies to consider longer-term epidemiological studies to look at other kinds of effects. For example, at the moment there is a big debate going on because both the International Agency for Research on Cancer, and the JECFA, are looking at the safety of aspartame and considering, for example, certain longer term epidemiological human studies. These bodies are talking about certain aspects of safety and looking at the very high levels of intake. Just to give you an example, at the moment the acceptable daily intake for aspartame is about 40 milligrams per kilogram per day, which is 2.8 grams for an average individual. That is about what you would get from 14 cans of a commonly consumed drink. That is a lot. We are saying maybe that is something that can be moderated. So these bodies talk about safe or maximal levels of intake of individual sweeteners. Here, in our recommendation, we are saying this practice of replacing sweeteners with sugars is not going to lead you to improved health gains, but you’re getting into an area of risk that even if it is within the agreed safety level, is something that you may want to avoid.

That is helpful context. So, what are the policy implications of all this?

Francesco: Some countries are already considering giving a warning on the content of non-sugar sweeteners in products. For example, what you have in Mexico. Some countries might decide to give the same warning if it is something that should be at least not encouraged. So that could be a way to do it. Other countries, while using nutrient profiling system, for example, for front of the pack labeling, or for marketing regulation, may decide to include sweeteners in the list of food that should be avoided. That is also happening, again, in South American countries. From our point of view, I think we would really like to give the message that the sweetness of food should be reduced. The same way we are having a dialogue with the food industry about reduction of the sugar content of food, particularly in foods that are intended for young children. We would really like to give a message that we should reduce the content of sweeteners. That is possible. And not relying entirely on sweeteners as an alternative for the reduction of sweetness or, reduction of calories, but really looking at a change in the strategies in peoples’ daily diets.

Thank you, I appreciate that. You know to the extent one person can be instructed in this, 11 or 12 years ago, or so, I stopped drinking diet beverages because of the very concerns we were talking about today. I just switched over to tea with no sweetener in it at all. I’m completely accustomed to that now and I find it very enjoyable. I also find the things that I used to drink to be completely oversweet now. So people can get calibrated at different levels of sweetness, can’t they?

Francesco: Absolutely. In a sense, this is the same discussion we’ve had with salt. I mean our capacity to adapt is there and it happens relatively quickly. People can enjoy foods without having that enormous level of sweetness. As we know, sometimes the excess presence of other salts, or sugar, or sweeteners, is some have a way to cover the lack of good taste from the other ingredients. So good foods and drinks can certainly happen without the unnecessary addition of this compound.