Story
Artificial & Non-nutritive Sweeteners
Last updated
In one pass Under a sugar-free label sit two completely different families. Not this — Diet soda is a healthy substitute — The World Health Organization (2023) conditionally advises against non-sugar sweeteners for weight control; the International Agency for Research on Cancer lists aspartame as Group 2B (possibly carcinogenic); in the Suez 2022 trial, saccharin and sucralose worsened blood-sugar responses.
Educational content, not medical advice — consult a clinician.
Story path
Chapter 1
Two kinds of sugar-free sweetener
Under a sugar-free label sit two completely different families. Tell them apart and you can predict whether a whole bar of sugar-free chocolate will send you to the toilet.
The first family fools the tongue. Its molecules fit the sweet-taste receptors on your tongue, and once a receptor is hooked it reports sweet to the brain. They are so intensely sweet that a pinch is enough. You cannot digest most of them, so they pass through the gut and leave unchanged. Aspartame is the exception: it is split into two amino acids, but the amount is so small that the calories can be ignored.
The second family is the sugar alcohols. They keep a sugar skeleton, and the small intestine absorbs only part of them. The rest reaches the large intestine, where it pulls water into the gut and is fermented by bacteria into gas. More water plus more gas means bloating and loose stools. One of them, xylitol, is highly toxic to dogs: if a dog eats some, contact a vet right away.
So low in calories does not mean no effect on the body. Sugar alcohols act right there in the gut. Whether sweeteners act somewhere else — on gut bacteria, on blood platelets, on the brain — is exactly what they are being studied for now.
The first family fools the tongue. Its molecules fit the sweet-taste receptors on your tongue, and once a receptor is hooked it reports sweet to the brain. They are so intensely sweet that a pinch is enough. You cannot digest most of them, so they pass through the gut and leave unchanged. Aspartame is the exception: it is split into two amino acids, but the amount is so small that the calories can be ignored.
The second family is the sugar alcohols. They keep a sugar skeleton, and the small intestine absorbs only part of them. The rest reaches the large intestine, where it pulls water into the gut and is fermented by bacteria into gas. More water plus more gas means bloating and loose stools. One of them, xylitol, is highly toxic to dogs: if a dog eats some, contact a vet right away.
So low in calories does not mean no effect on the body. Sugar alcohols act right there in the gut. Whether sweeteners act somewhere else — on gut bacteria, on blood platelets, on the brain — is exactly what they are being studied for now.
Mechanism · The six non-nutritive sweeteners
Under the everyday phrase artificial sweetener, this story sorts 8 families. Their chemistry, their fate in the body, and their safety evidence differ widely, and treating them as one class is where many misunderstandings start.First, the 6 non-nutritive sweeteners, which supply almost no calories. One line each:
Aspartame: about 200 times as sweet as sucrose, and taken apart in the small intestine. People with phenylketonuria (PKU), an inherited inability to process phenylalanine, must avoid it. Many zero diet sodas use it. In 2023 the International Agency for Research on Cancer (IARC), part of the World Health Organization, placed it in Group 2B, meaning possibly carcinogenic. That grade describes how certain the evidence is, not how large the risk is.Sucralose: about 600 times as sweet as sucrose. Most of it is not absorbed, and it is heat-stable enough to bake with. The newer question mark is a molecule called sucralose-6-acetate. Part of it is an impurity left from manufacturing, and part can be made from sucralose by gut bacteria. An in-vitro screening study reported that it is genotoxic (Schiffman 2023). That is a cell experiment, not data on human exposure.Saccharin: the oldest artificial sweetener, about 300–400 times as sweet as sucrose. It is not metabolized and leaves unchanged. The old warning about bladder cancer in rats turned out to rest on a mechanism that works only in rats, and the United States dropped the warning label around 2001.Acesulfame-K: about 200 times as sweet as sucrose and not metabolized. It is usually blended with other sweeteners, because the blend tastes closer to sucrose.Stevia: extracted from the South American stevia plant, about 200–300 times as sweet as sucrose. Your own digestive enzymes cannot split it, but bacteria in the large intestine do. The freed part is absorbed, processed by the liver, and leaves in urine. It is marketed as natural, but natural does not mean safe: the US FDA accepts only the high-purity extract.Monk fruit: a traditional Chinese source of sweetness. Its sweet compounds, the mogrosides, are about 200 times as sweet as sucrose and are also broken down mainly by gut bacteria. In 2010 the US FDA raised no objection to its use.
The sweetness multiple is itself a mechanism
A sweet-taste receptor does not care how many grams you add. It responds to how many molecules lock into its pocket. The molecules above bind far more tightly than sucrose, so a tiny amount fills the receptors. The sweetness you taste is at full scale, while the material that actually enters your body is a trace. Their calories are near zero not because the molecules hold no energy, but because the dose is so small.
Aspartame is the only one your own digestive enzymes take apart
In the small intestine it is cut into two amino acids, aspartate and phenylalanine, plus a little methanol. Both amino acids already occur in food protein, and a bite of egg gives you far more of them. It still counts as zero-calorie because the dose is small and the calories from those amino acids can be ignored.
So what matters is not calories but who that freed phenylalanine lands on. For almost everyone it is an ordinary amino acid that follows the same path as the ones in food protein. For the few people born unable to process it, it builds up in the blood. That is also why the label says "contains phenylalanine" rather than "contains aspartame" — explained in detail in the later chapter on aspartame's cancer grade.
The sweetener umbrella also covers 2 families that carry some calories and have real metabolic effects: sugar alcohols and allulose. Lumping them in with the 6 above is a different misunderstanding.
Mechanism · How sugar alcohols and allulose differ
Family 7: sugar alcohols. Unlike the first 6, they carry some calories and have metabolic effects:Erythritol: 4 carbons, about 0.24 kcal per gram, about 70% as sweet as sucrose. Its molecule is small, so most of it is absorbed into the blood in the small intestine and leaves unchanged in urine. Little reaches the large intestine, so it is barely fermented and small amounts rarely cause gas. That same advantage is the source of its newer problem, covered in the later chapter on erythritol and blood platelets.Xylitol: 5 carbons, about 2.4 kcal per gram. The mouth bacteria that cause tooth decay cannot use it, so it is often added to chewing gum and toothpaste. It is highly toxic to dogs.Sorbitol, maltitol, isomalt, mannitol: poorly absorbed in the small intestine. The rest reaches the large intestine and is fermented by bacteria, so bloating and osmotic diarrhea — the dissolved sugar alcohol pulls water into the gut — are common.
Why sugar alcohols give people diarrhea
A sugar alcohol that is not absorbed stays in the gut like a handful of salt dissolved in water. It draws water out of the gut wall into the gut. In the large intestine, bacteria then ferment it and make gas. More water plus more gas means bloating and loose stools. Tolerance varies from person to person, and 10–50 g can already cause symptoms. Sugar-free chocolate and candy often use maltitol, and eating 100 g at once can cause diarrhea, where the same amount of sucrose would not. Erythritol is the exception here, because most of it is absorbed in the small intestine, but very large amounts can still upset the stomach.
Family 8: allulose, a newer rare sugar:
It is an isomer of fructose — the same atoms, arranged differently. It is about 70% as sweet as sucrose and provides about 0.4 kcal per gram. After absorption it largely stays out of the cell's main energy pathway, the tricarboxylic acid () cycle, and most of it leaves in urine.Since 2019 the US FDA has allowed it to be left out of total sugars on the nutrition label.So far it does not appear to raise blood glucose, but long-term data are sparse and it is expensive.
Why these two families get their own section
"I don't eat sugar" while drinking a lot of aspartame diet soda is not the same as "I use xylitol instead of sugar". The first brings almost no calories. The second brings about 2.4 kcal per gram, and it ferments.Diarrhea after too much sugar-free chocolate is a direct result of the physical chemistry above.Each family has its own metabolic effects and its own safety evidence, so the questions of cancer grading, gut bacteria, and platelets all have to be taken one family at a time.
In practice · What common sugar-free products contain
What goes into common "0 sugar" and "sugar-free" drinks and foods in China (recipes change, so the ingredient list on the item in your hand is what counts):Sodas and tea drinks
GenkiForest sparkling water: erythritol plus sucralose.Zero-sugar colas such as Coke Zero: mostly aspartame plus acesulfame-K.Oriental Leaf (unsweetened): no sweetener at all; it is just tea.Nongfu Spring NFC juice: real juice with no added sugar, but juice naturally contains sugar, so it is not "0 sugar".
Milk tea and coffee labeled "light" or "sugar-free"
A common approach is to cut the sugar by half (50% less sugar) and make up the sweetness with aspartame, acesulfame-K, or sucralose. The tapioca pearls, coconut jelly, and fruit jam are themselves high in sugar.Sugar-free milk tea is not low-calorie; one cup can still have 300–400 kcal.
Packaged snacks labeled "0 sugar"
Sugar-free cookies and cakes: often maltitol plus sucralose, but the flour and fat are still there, so the calories from carbohydrate and fat have not gone down.Sugar-free chocolate: mostly maltitol, which causes diarrhea if you eat a lot.
Children and infants
Infant formula: Chinese and international standards do not allow added sweeteners.Children's vitamin gummies: most contain sucrose or glucose syrup and are not sugar-free."0 sugar" drinks for school-age children: the same as the adult versions, and safety data in children are sparse. WHO's 2023 recommendation not to use non-sugar sweeteners for weight control applies to everyone except people who already have diabetes. It is not a rule only for children, and in WHO's own grading it is a conditional recommendation.
People with diabetes and people trying to lose weight
"Diabetic" cookies: often maltitol or sorbitol. Blood glucose rises more slowly, but the calories are not much lower, and eating a lot causes diarrhea.How much weight does swapping sugar for sweetener take off? In the Toews 2019 published in the BMJ, body mass index () was only about 0.6 lower. That figure rests on just two small studies with 174 people, and the certainty of evidence was rated low to very low. The problem is not that the result was "not significant"; it is that we have little confidence in the number itself.
The point to remember: 0 sugar is not 0 calories, and even less is it healthy. Judge a packaged food by reading the ingredient list, total calories, total carbohydrate, protein, and fat together, not by trusting a single sugar-free label.
Chapter 2
What aspartame's cancer rating means
That headline traveled as aspartame causes cancer, but the International Agency for Research on Cancer (IARC) did not say that.
IARC answers a yes-or-no question: can this thing cause cancer at all. It placed aspartame in Group 2B, meaning possibly carcinogenic, on the basis of limited evidence in people. The grade describes how certain the evidence is. It does not say whether the amount you drink is dangerous. It is like being told that the river downstairs can drown someone: true, but it does not say whether you are standing on the bank or already in the water.
How deep you are standing is worked out by a different committee that deals with dose, JECFA. The same year, it reconfirmed aspartame's acceptable daily intake (ADI): 40 mg per kilogram of body weight.
Aspartame's path through your body is short. In the small intestine it is cut into two amino acids plus a little methanol, and it does not enter the blood as aspartame. The real concern is one of those amino acids, phenylalanine, when it lands on the few people born unable to process it.
IARC answers a yes-or-no question: can this thing cause cancer at all. It placed aspartame in Group 2B, meaning possibly carcinogenic, on the basis of limited evidence in people. The grade describes how certain the evidence is. It does not say whether the amount you drink is dangerous. It is like being told that the river downstairs can drown someone: true, but it does not say whether you are standing on the bank or already in the water.
How deep you are standing is worked out by a different committee that deals with dose, JECFA. The same year, it reconfirmed aspartame's acceptable daily intake (ADI): 40 mg per kilogram of body weight.
Aspartame's path through your body is short. In the small intestine it is cut into two amino acids plus a little methanol, and it does not enter the blood as aspartame. The real concern is one of those amino acids, phenylalanine, when it lands on the few people born unable to process it.
Evidence · Hazard grade versus safe dose
In July 2023, the International Agency for Research on Cancer (IARC), part of the World Health Organization, placed aspartame in Group 2B, meaning possibly carcinogenic. The headline was widely misread, so it is worth taking apart.What IARC's groups are
Group 1 (carcinogenic to humans): sufficient evidence in people. Examples: tobacco, alcohol, processed meat, solar radiation, asbestos, formaldehyde.Group 2A (probably carcinogenic): usually limited evidence in people but stronger evidence from animal experiments or mechanisms. Examples: red meat, night-shift work, acrylamide, very hot drinks above 65°C.Group 2B (possibly carcinogenic): limited evidence in people, or sufficient evidence only in animal experiments. Examples: gasoline, whole-leaf aloe extract, and aspartame. For aspartame this time, the human, animal, and mechanistic evidence were all rated limited.Group 3 (not classifiable): not enough evidence to reach a conclusion. A large share of everything evaluated sits here, coffee itself included.The former Group 4 (probably not carcinogenic): abolished when the evaluation handbook was revised in 2019. Its single entry at the time, caprolactam, moved to Group 3.
The coffee and very-hot-drink lines are often stated backwards. The same 2016 IARC evaluation did two things. It moved coffee down from Group 2B to Group 3, because it found no sufficient evidence that coffee itself causes cancer. At the same time it placed drinking very hot beverages above 65°C in Group 2A. The problem is the temperature, not the coffee, and very hot drinks sit one grade above aspartame. Using "coffee causes cancer" as the example for Group 2B teaches this whole lesson backwards.
The table ranks certainty, not danger
Processed meat and asbestos sharing a group does not mean eating ham equals inhaling asbestos. The groups rank how certain the evidence is: the higher the group, the harder it is to overturn the statement this thing can cause cancer. It says nothing about what happens at the amount you actually eat.
Dose belongs to a different committee
Judging how large the risk is at real intakes is the job of the Joint FAO/WHO Expert Committee on Food Additives (JECFA).JECFA's parallel evaluation in 2023 reaffirmed aspartame's acceptable daily intake (ADI) at 40 mg per kilogram of body weight.By WHO's own figure, a can of diet soda contains 200 or 300 mg of aspartame. For a 70 kg adult the ADI is 2,800 mg, so with no aspartame from any other food they would need to drink more than 9–14 cans a day to cross that line. The lighter the person, the fewer cans it takes.Almost everyone's intake is far below the ADI.
Splitting the two jobs between two bodies is deliberate. One asks only whether there is a signal; the other asks how big that signal is at real intakes. Mixing them up while reading the news gives you a conclusion that is both frightening and useless: you do not know what to change, or how far is far enough.
iarc-2016-coffee
Evidence · Why the human evidence is only limited
What IARC based Group 2B onHuman evidence: mainly a signal for hepatocellular carcinoma, the most common kind of liver cancer, in a few large observational cohorts. It was rated limited: not enough to conclude, not enough to rule out. Cohorts such as France's NutriNet-Santé have also reported associations between artificial sweeteners and cancer, and those are observational too.Animal evidence: including the rat experiments of Italy's Ramazzini Institute, rated limited. The methods of those experiments have also been criticized.Mechanistic evidence: also limited.Confounding is hard to rule out: people with diabetes or obesity already have a higher cancer risk, and they are also the people who drink the most sweeteners.
What this means for you
1–2 cans of diet soda a day: far below the ADI. The epidemiological signal is weak, so the risk looks very small, but it cannot be called zero.5–10 cans a day: by WHO's per-can figure, 10 cans at 300 mg each already exceed the ADI of a 70 kg adult. Even below that line, drinking this much soda every day is worth rethinking in its own right.Children and pregnant women: the data are sparse, so extra caution is reasonable.Switching to water or tea: the safer choice at any level of intake.
How to read diet soda and cancer
The headline aspartame causes cancer is technically wrong. IARC said possibly, and it was only ranking how certain the evidence is.A more accurate statement: the evidence is weak and the estimated real-world risk is small, but it has not been shown to be completely safe.Is it better than the sugary version? For people already in the habit of drinking soda, the diet version may be slightly better, since it drops sugar's direct effects on blood glucose and weight. But the best choice is not drinking soda.
Risk, size of harm, actual exposure, and whether to change what you do are four different things. Science coverage often merges them into one.
The direction of is worth walking through slowly
The causal arrow here is easy to read backwards. It may not be sweeteners that give people diabetes or obesity. Part of it is more likely the reverse: people who already have high blood glucose or high body weight are the ones who switch to diet drinks, and this group already has a higher cancer risk than the general population. So in the association people who drink diet drinks get slightly more cancer, part of it is illness first, then the bottle.
To strip that part out you need random assignment: put people into drink and no-drink groups at random, so the two groups average out on everything else. But the endpoint is cancer decades later. Almost nobody can afford that trial, and nobody can wait for it. So this grade of evidence will probably stay at limited for a long time — not because scientists are lazy, but because the shape of the question does not allow harder evidence.
Knowing this, you can judge the next story of this kind yourself. First ask whether it is an association or a randomized comparison. Then ask how far its exposure is from your everyday life.
Safety · Who really needs to avoid aspartame
Who really needs to avoid aspartame, or be careful with it1. Phenylketonuria (PKU)
An autosomal recessive inherited disease: people are born without enough of the liver enzyme that converts phenylalanine into tyrosine, phenylalanine hydroxylase (PAH). It is uncommon, and newborn screening checks for it in China and in many other countries.People with PKU cannot clear phenylalanine, so it builds up in the blood. At high levels it is toxic to the developing brain and damages intellectual development. The damage is serious, but early diagnosis and early dietary control can prevent it.About half of the aspartame molecule is phenylalanine, so people with PKU must avoid it.Regulations require foods containing aspartame to say contains phenylalanine.Because of newborn screening, people with PKU usually know they have it.
2. Pregnancy
The shared position of the US FDA, Europe's EFSA, and WHO is that aspartame is safe within the ADI.A few observational studies have raised questions. In Canada's CHILD birth cohort (Azad 2016, JAMA Pediatrics, N = 3,033 mother–infant pairs), mothers who drank artificially sweetened drinks daily during pregnancy had babies with a higher body mass index () at 1 year, and about 2 times the risk of overweight compared with mothers who did not drink them. The US Project Viva birth cohort (Plows 2022) also reported that higher sweetener intake in pregnancy went with higher BMI and body fat in childhood. These are associations and do not show that the sweeteners caused anything.The current consensus is no ban, but drink less, with water and unsweetened drinks as the daily default.
3. Depression and mood
A few small studies have suggested a link between aspartame and worse mood or depressive symptoms. One idea is that large amounts of phenylalanine disturb the brain's pathway from tyrosine to dopamine.The evidence is weak and does not justify telling everyone to stop. If you suspect a link in yourself, stopping for a while to see does no harm, but it is not a substitute for proper treatment of depression.
4. Migraine: this one was tested double-blind and did not reproduce
Aspartame triggering headaches is the most widely repeated claim, but it has been tested head-on and did not hold up. Schiffman's 1987 trial in NEJM recruited exactly the people who said they got repeated headaches after aspartame — 40 of them — and gave them aspartame or placebo in a double-blind crossover. Headaches followed aspartame 35% of the time and placebo 45% of the time, with no significant difference. The authors concluded that in this group aspartame was no more likely than placebo to cause headaches.That does not mean your headaches are imaginary. It means the trigger is probably not this molecule. Common migraine triggers are sleep, the menstrual cycle, dehydration, a skipped meal, and a sudden drop in stress. People often reach for a diet drink at exactly those moments, so the drink takes the blame.A headache diary can certainly log aspartame, but the point is to read your own data, not to confirm a claim that has already been tested and failed. If you really want to know, stop for two weeks and then try it again, instead of going on impressions.
Erythritol: the other one people ask about in the same breath
It has nothing to do with aspartame, but the two are often asked about together. Its new question is not cancer but the heart and blood vessels. In an observational study from the Cleveland Clinic (Witkowski 2023, Hazen's group), the quarter of people with the highest blood erythritol had about 2 times the 3-year risk of major adverse cardiovascular events (: death, heart attack, or stroke) of the lowest quarter, and in experiments erythritol made platelets easier to activate. This is an association and causation has not been established; the evidence is laid out in the later chapter on erythritol and platelets.
Connecting the PKU and mood threads
The enzyme missing in PKU does exactly one job: turning phenylalanine into tyrosine. The pathway in the mood section starts from that same tyrosine and, step by step, becomes dopamine. So missing this one enzyme causes two things at once. Upstream, phenylalanine backs up in the blood and keeps rising. Downstream, the tyrosine it should have become runs short. The damage to the nervous system is squeezed out of both ends together, not simply one substance being toxic.
That also explains the wording on the label: the law requires contains phenylalanine, not contains aspartame. For someone with PKU the danger was never the sweet molecule; it is the half released when that molecule is cut in the small intestine. By the same logic, phenylalanine from anywhere else — a piece of meat, a glass of milk — is just as risky. What they manage is their total phenylalanine for the day, not avoidance of one brand.
schiffman-1987-nejm-aspartame-headacheplows-2022-nns-pregnancy-child-bmi
Chapter 3
How sweeteners change gut bacteria
Whatever you do not digest, something else will deal with.
Not absorbed by the body has a second half. The sweetener travels to the large intestine and lands, intact, in front of tens of billions of bacteria. For you it has zero calories; for some of those bacteria it is something to use or something to cope with.
The bacteria that can use it or tolerate it multiply, others shrink back, and the guest list in your gut changes. Those guests do two jobs that matter for blood glucose. They ferment fiber into short-chain fatty acids, and they remodel the bile acids that flow down from the liver. Both kinds of molecule cross the gut wall into the blood and shift your insulin sensitivity — how high your blood glucose climbs after the same bowl of rice.
So this drink has no sugar does not lead to it does not affect your blood glucose. Two studies from Israel's Weizmann Institute support this chain: in mice they isolated the role of the gut bacteria, and in people they saw an effect that varied from person to person.
Not absorbed by the body has a second half. The sweetener travels to the large intestine and lands, intact, in front of tens of billions of bacteria. For you it has zero calories; for some of those bacteria it is something to use or something to cope with.
The bacteria that can use it or tolerate it multiply, others shrink back, and the guest list in your gut changes. Those guests do two jobs that matter for blood glucose. They ferment fiber into short-chain fatty acids, and they remodel the bile acids that flow down from the liver. Both kinds of molecule cross the gut wall into the blood and shift your insulin sensitivity — how high your blood glucose climbs after the same bowl of rice.
So this drink has no sugar does not lead to it does not affect your blood glucose. Two studies from Israel's Weizmann Institute support this chain: in mice they isolated the role of the gut bacteria, and in people they saw an effect that varied from person to person.
Evidence · How two studies pointed to gut bacteria
Artificial sweeteners, gut bacteria, and blood glucose have been among the most debated research lines of the past 10 years.Suez 2014 (Nature, Weizmann Institute, Israel)
Mouse experiments: mice given commercial sucralose, aspartame, or saccharin formulations in their water showed changed gut bacteria and worse glucose tolerance — after a dose of sugar, their blood glucose took longer to come back down. Wiping out the bacteria with antibiotics made the effect disappear.A small human study (N = 7 healthy people who did not normally use sweeteners): after about a week of saccharin, 4 of the 7 had worse glucose tolerance, and their gut bacteria changed too; the other 3 barely changed. The study was tiny and had no control group.The key step: gut bacteria from the participants whose glucose tolerance worsened were transplanted into germ-free mice, and those mice developed worse glucose tolerance as well. In mice, this supports the bacteria being the link in the middle.
Suez 2022 (Cell): a in N = 120 healthy adults
6 groups of 20 people each: 4 sweetener groups (saccharin, sucralose, aspartame, stevia) and 2 control groups — one given only the glucose carrier in the same kind of sachet, one given nothing. Supplements ran for 2 weeks, all at doses below the ADI.Each of the four sweeteners changed, in its own way, the bacteria in stool and in the mouth, and the metabolites in the blood.The saccharin and sucralose groups, taken as groups, had significantly worse blood-glucose responses. The aspartame and stevia groups showed no such change at the group level.Individual differences were large: on the same dose, some people clearly worsened and some barely changed. The researchers transplanted the bacteria of the strongest and weakest responders into germ-free mice, and the mice's glucose responses largely mirrored those of their human donors.
The proposed mechanism
Sweeteners the body does not absorb come into direct contact with gut bacteria.In the 2014 mouse experiments, some bacteria in the order Bacteroidales increased and some in the order Clostridiales decreased. Once the community shifts, the handling of short-chain fatty acids and bile acids shifts too, which in turn changes insulin sensitivity and the glucose response. The second half of this chain is mostly inference and has not been measured step by step in people. The short-chain fatty acid pathway itself is covered in more depth in Carbohydrates and fiber.Sucralose carries chlorine atoms, and one idea is that this makes its effect on the bacteria stronger.
A glycemic index () of zero is not the same as neutral for blood glucose
The classic nutritional reasoning was: no sugar, so no effect on blood glucose.The two Suez studies suggest that sweeteners can affect blood glucose indirectly, through the gut bacteria, in some people.Sweeteners treated as neutral for sugar metabolism may be less neutral than assumed; in whom, and by how much, is not yet clear.
What it means for medicine and policy
WHO's 2023 guideline recommends against using non-sugar sweeteners to control body weight or to lower the risk of noncommunicable diseases. The systematic review behind it included 283 studies and found no long-term benefit for reducing body fat.Three qualifiers must stay attached to this recommendation. First, in WHO's own grading it is a conditional recommendation, not a strong one, because the associations seen in the studies may be by the participants' own health. Second, it applies to everyone except people who already have diabetes. Third, it explicitly does not cover sugar alcohols: erythritol and xylitol are outside this guideline, and using it to judge them oversteps it.WHO did not ban any sweetener. It only recommends against using them for these two purposes, and how to regulate them is left to each country.
Why moving the bacteria into germ-free mice is the critical step
Seeing that people whose blood glucose worsened also had different gut bacteria leaves two equally plausible explanations. The bacteria made blood glucose worse, or worse blood glucose changed the bacteria. From human data alone, the two look identical.
Germ-free mice separate them. These mice have never met a bacterium since birth, and their blood glucose starts out normal. The researchers moved exactly one thing into them: the bacteria. If the problem moves house along with the bacteria, the arrow points from bacteria to blood glucose. That is a rare bridge between association and causation, and it is why this research line deserves to be taken seriously. Its limits need saying too: the bridge was built in mice, while the human trial lasted only 2 weeks and was not large.
Why sucralose stirs things up more (one explanation)
Its backbone is sucrose, with a few positions swapped for chlorine atoms. Because the backbone is still there, gut bacteria still somewhat recognize it. The chlorines stop human digestive enzymes from cutting it, so it arrives in the large intestine intact and is handed to the bacteria. A molecule that looks like food but does not belong to you shows up in front of the same residents every day. That is the intuitive version of the idea that chlorine may make the disturbance stronger, and it has not itself been tested directly.
Evidence · Which is worse, sweetener or sugar
Sweeteners or sugar: which is worse?Sugar (sucrose, high-fructose corn syrup or ) at high intake: the evidence is stronger. In randomized trials, eating more sugar raises body weight. In large cohorts, people who drink more sugary drinks have more type 2 diabetes, fatty liver, and cardiovascular disease.Artificial sweeteners at high intake: they may disturb the gut bacteria, sugar metabolism, and appetite regulation, with evidence mostly from small trials and animal experiments.Neither is good, but the evidence against sugar is stronger and the harm is larger.The best thing to switch to is water, tea, or black coffee, not the diet version in place of the regular one.
Compared with real food
Whole fruit contains natural sugar, but it also brings fiber, polyphenols, and vitamins, and on balance it counts in your favor."Natural" sugars such as honey and maple syrup are still sugar, with no clear advantage.Natural versus artificial is the wrong axis. Whole food versus industrial product is the real one.
Why whole food versus industrial product is the real axis
The sugar in an apple and the sugar in a can of soda are the same molecule, but they reach the small intestine at completely different speeds. The apple's sugar is locked inside a scaffold of cell walls and fiber. You have to chew it first, and then digestive enzymes release it a little at a time. The sugar in a drink is already dissolved in water and can pass through the moment it reaches the stomach. The same amount of sugar arrives as a trickle in one case and as an opened floodgate in the other, and your insulin has to handle two completely different curves.
The natural versus artificial axis cannot tell the two apart: honey is natural, yet it stands on the floodgate side. What really separates them is whether the food still carries its original structure.
Mechanism · What the brain does with sweet but no sugar
Sweeteners have a second route that runs through the brain, not metabolismExpectation first, metabolic response second: the cephalic phase
Before sugar even reaches your mouth, seeing, smelling, or thinking about it makes the brain release a little insulin and some gut hormones ahead of time, to prepare for digestion. This is called the cephalic phase response, and it was described decades ago.One hypothesis: when you drink a diet drink, the sweetness triggers the same expectation, but the sugar never arrives. After many repeats, the brain may adjust how it responds to sweetness. Measurements of this step in people have been inconsistent.
The long-term hypothesis (Yang 2010, a review)
Repeated sweetness without sugar may blunt the brain's reward circuit, so that real sugar no longer satisfies as much, and it may also stir up cravings for sugar. This hypothesis is inferred from animal experiments and mechanisms and has not been confirmed in people.The Mattes 2009 review paints a different picture. When sweeteners replace sugar, people usually make up only part of the calories saved, so compensation is incomplete. Sweeteners may increase hunger only when they are added to something with no calories; eaten with other food, that effect is not seen. And most of the proposed mechanisms by which sweeteners would promote eating are not supported by the evidence available. The review also notes that there is no evidence yet on whether sweeteners help with long-term weight control.Pooled randomized trials: swapping sugar for sweetener does make people slightly lighter in the short term, but by less than is usually claimed. Miller 2014 (AJCN) pooled the randomized trials and found 0.80 kg less body weight (95% −1.17 to −0.43). Toews 2019 (BMJ) reported a body mass index () about 0.6 lower (95% confidence interval −1.19 to −0.01 — note the unit is BMI, not kilograms), and that figure rests on just 2 small studies with 174 people, with the certainty of evidence rated low to very low.So this evidence is soft too. It supports the direction may be right, not how many kilograms you will lose.One way to reconcile them: a short-term swap really does remove calories. Whether that holds long-term depends on whether the calories come back elsewhere, and no long-term trial has answered that yet.
The special question for children
Food experiences in childhood influence taste preferences later in life, a point that comes mainly from observational research and experience.One concern is that frequent sweet drinks make natural foods seem not sweet enough, so children like fruit and vegetables less. This is a guess and has not been measured directly."Sugar-free" drinks for kids look like the better option, but they reinforce the habit that drinks are supposed to be sweet.The general advice from pediatric and nutrition bodies is water and plain milk as children's everyday drinks. When WHO released its 2023 guideline, it also said that the overall sweetness of the diet should be lowered, starting early in life.
The idea of sugar addiction
Sugar does activate the brain's dopamine reward pathway, and some animal experiments have shown addiction-like behavior.In people, however, sugar addiction is not an established diagnosis. It is closer to habit and reward learning than to physical dependence.Switching to diet drinks does not solve this: the core is expecting sweetness, and diet drinks meet that expectation just as well.A common suggestion is to cut all sweet input, sugar and sweeteners together, for a period, so your taste recalibrates. The advice is widespread, but it has barely been tested in people, so do not expect a set number of weeks to make fruit taste sweeter.
How to use them in practice
Swapping the regular version for the diet one is a tool for cutting calories in the short term, not a long-term solution.The long-term goal is to lower the overall sweetness of your drinks and snacks.It is a gradual process; there is no need to quit sugar overnight.
yang-2010-sweetener-brainmiller-2014-ajcn-lcs-weight
Chapter 4
Erythritol and blood platelets
Erythritol used to look like the free lunch among sweeteners: closest to sugar in taste, no rise in blood glucose, no food for the bacteria that cause tooth decay, and rarely any gut trouble.
The reason it spares the gut is concrete. Its molecule is the smallest of the common sugar alcohols. Before it reaches the large intestine, most of it has already been absorbed into the blood in the small intestine and then leaves in urine, so the bacteria in the large intestine never get it. Other sugar alcohols stay in the gut, get fermented, and pull water in; erythritol mostly does not.
The new problem grows out of that very advantage. Because erythritol enters the blood whole, it meets the platelets circulating there. In experiments, as the erythritol concentration rises, platelets become easier to activate and more likely to clump, and platelet clumping is the first step of a blood clot. So gentle on the gut and possibly not gentle on blood vessels are two sides of the same fact.
This chain has reached an association in people plus a plausible mechanism. It has not reached proof in a randomized trial. For people with heart disease or a high risk of clots, cutting back on it now is reasonable.
The reason it spares the gut is concrete. Its molecule is the smallest of the common sugar alcohols. Before it reaches the large intestine, most of it has already been absorbed into the blood in the small intestine and then leaves in urine, so the bacteria in the large intestine never get it. Other sugar alcohols stay in the gut, get fermented, and pull water in; erythritol mostly does not.
The new problem grows out of that very advantage. Because erythritol enters the blood whole, it meets the platelets circulating there. In experiments, as the erythritol concentration rises, platelets become easier to activate and more likely to clump, and platelet clumping is the first step of a blood clot. So gentle on the gut and possibly not gentle on blood vessels are two sides of the same fact.
This chain has reached an association in people plus a plausible mechanism. It has not reached proof in a randomized trial. For people with heart disease or a high risk of clots, cutting back on it now is reasonable.
Evidence · What the erythritol study measured
Erythritol was long treated as the perfect sweetener. A 2023 study from the Cleveland Clinic shook that judgment.Where the "perfect sweetener" reputation came from
About 0.24 kcal per gram, almost no caloriesNo rise in blood glucose, because it does not enter glucose metabolismMostly absorbed in the small intestine, so rarely fermented and unlikely to cause diarrhea unless you eat a great dealMouth bacteria cannot use it, so it does not promote tooth decayFound naturally in fruit and made industrially by fermentationAbout 70% as sweet as sucrose, with the taste closest to sugar
What the study did (Witkowski 2023, Nature Medicine, Stan Hazen's group)
Four layers of evidence: population cohorts, test-tube experiments, mouse experiments, and a drinking test in healthy people
1. Population: the participants were stable patients having a planned heart evaluation, so their cardiovascular risk was already above average; the three cohorts together held more than 4,000 people. Comparing the highest quarter of blood erythritol with the lowest quarter, the 3-year risk of major adverse cardiovascular events (: death, nonfatal heart attack, or stroke) was about 2 times higher. In the US validation cohort (n = 2,149) the (HR) was 1.80 (95% 1.18–2.77); in the European validation cohort (n = 833) it was 2.21 (95% confidence interval 1.20–4.07). This is an association observed after adjusting for other risk factors.
2. Test tube: at concentrations found in human blood, erythritol made platelets easier to activate.
3. Mice: when erythritol in the mice's blood was raised to that level (by injection), clots formed faster after the carotid artery was injured.
4. Healthy people: 8 healthy volunteers drank a beverage containing 30 g of erythritol — an amount the authors compare to one can of a commercial erythritol drink or one pint of keto ice cream. Their blood levels rose about 1,000 times above baseline and stayed clearly raised for more than 2 days, far above the threshold that made platelets more active in the test-tube experiments.
Why this matters
The core argument 0 calories, 0 , natural, therefore safe has been shaken.Many keto, diabetic, and weight-loss foods, and many 0 sugar drinks in China (such as GenkiForest sparkling water), use erythritol.It is still permitted, and the researchers' own conclusion is that its long-term safety needs dedicated study.
Why the blood-level finding is more persuasive than the hazard ratio
Looking at the association alone, you can always push back: people with high blood erythritol might already be sicker, and might also eat more sugar-free processed food. That objection is hard to knock down.
But when a healthy person drinks an ordinary serving and the blood level shoots up and stays up for days, that says something else. The amount people eat in daily life is enough to push blood levels into the range where problems appeared in the study. Dose and exposure line up, and the it is just an association defense becomes much harder, because it swaps people with high blood levels are a special group for an ordinary person reaches that level after one serving.
Add the test-tube and mouse layers, and the three pieces connect: the level rises, platelets clump more easily, clots form faster. The missing link is randomly assigning healthy people to drink it or not and following such as heart attack and stroke. Nobody has done that yet.
Safety · Who should cut back on erythritol now
Limits of the study, stated plainlyObservational, so it shows association only: people at high cardiovascular risk already have more , and they may also eat more 0 sugar foods. That is . The authors themselves write that such studies can show association but not causation, and that factors such as diet may not have been fully adjusted for.Blood erythritol does not all come from food: the body makes small amounts itself. The authors suspect the blood levels in the cohorts combine what was eaten with what the body made. When the US cohort was enrolled, erythritol had not yet spread widely through processed food, so fasting levels then probably reflected the body's own production.Measured once: each person in the cohorts had a single fasting measurement, taken at enrollment.No : there is still no trial that gives healthy people erythritol or a placebo and then tracks heart attacks and strokes.The mouse mechanism experiments and the roughly 1,000-fold rise in blood levels in healthy people weaken the it is just an association defense: the dose fits, and the mechanism makes sense.The findings still need to be replicated by independent groups.
What to do now
A history of cardiovascular disease or a high risk of clots (atrial fibrillation, a past stroke, taking anticoagulants): cutting back on erythritol is reasonable.Healthy people: small amounts (once or twice a day) probably carry little risk, but it can no longer be called completely safe.Alternatives: allulose, monk fruit, stevia, a very small amount of sucrose, and whole foods.Using no sweetener at all is still the best choice: water, tea, black coffee.
Xylitol shows a similar signal
The same group's 2024 study in the European Heart Journal (Witkowski 2024): people with higher blood xylitol also had a higher 3-year risk of MACE, again an observational association. In experiments xylitol likewise made platelets more active, and healthy people's platelet reactivity rose after they drank a xylitol-sweetened beverage.Xylitol is common in diabetic foods, anti-cavity chewing gum, and toothpaste.Xylitol is acutely lethal to dogs (a dog releases a flood of insulin, causing severe low blood sugar and possibly acute liver failure). If you have a dog, never let it get at any food containing xylitol.
What this means for the old consensus
The long-standing consensus that sugar alcohols are simply safe is being revised by new research.There is no need to panic: the risk signal is real, but on current data the absolute risk is still small.If you have been assuming zero-calorie erythritol is a harmless substitute, that view no longer holds.
Why xylitol is an emergency for dogs but not for people
This is not a vague dogs are more fragile than people story; it is a specific case of mistaken identity. Once xylitol is in a dog's blood, the dog's pancreas treats it as a sugar signal that needs handling right away and releases a flood of insulin. The human pancreas does not respond to it this way. With that much insulin, glucose is driven out of the blood into cells in bulk, blood sugar is dragged down to a dangerous low within a short time, and acute liver failure may follow.
So a small piece of sugar-free gum dropped on the floor and picked up by a dog is a real emergency, not a little won't hurt. If your dog eats some, contact a vet at once; do not wait for symptoms. If you have a dog at home, keep xylitol toothpaste and gum where the dog cannot reach them.
Myth · Does natural mean safe
Natural does not mean safe, and artificial does not mean dangerousThe erythritol case, and the whole sweetener debate, point to the same way of judging.
Common ways "natural" gets mistaken for "safe"
"Honey is a natural sugar, so it beats white sugar": honey and white sugar are both sugar at heart, with similar calories and similar effects on blood glucose."Brown sugar has minerals, so it beats white sugar": the amount of minerals is tiny and does not add up to a nutritional benefit."Maple syrup and coconut sugar are natural": they are all sugar, and too much of any of them does harm."Stevia and monk fruit are natural": they are sweeteners, not sugars, but being natural does not prove they are safe.
Some natural things really are dangerous
Compounds that release cyanide (cyanogenic glycosides): bitter almonds, raw cassava, ginkgo nuts.Natural toxins: poisonous mushrooms, and the leaves and seeds of some plants.Compounds with hormone-like or drug-like activity: soy isoflavones (a type of plant estrogen), red yeast rice (its monacolin K is the same molecule as the cholesterol-lowering statin lovastatin), and some herbs.Natural carcinogens: safrole (from sassafras root), aflatoxins (mold toxins in spoiled grains and peanuts), aristolochic acid (in some herbal remedies).
Some artificial things really have advantages
Precise dosing: medicines, and nutrient fortification (iodized salt, flour fortified with folic acid).Reproducible and regulated.Less contamination: modern food safety has greatly reduced natural hazards such as molds and parasites.
The axes worth looking at
Whole food or industrial product, whether natural or artificial.Strength of evidence: which one has more studies, whether they looked at hard outcomes, and whether they followed people long enough.Dose and accumulation: whether something acts as a poison depends on the dose.
Back to sweeteners
Stevia (natural) and sucralose (artificial): do not pick stevia just because it is natural; look at the evidence, your own response, and the price.Honey (natural) and erythritol (found in nature, made industrially): both can be called natural, yet their metabolism and clinical evidence are completely different.Whole fruit versus any sweetener: whole fruit is the default best choice, because it is more than a carrier of sweetness.
So judging any sweetener comes back to the same habit: do not decide by the label; look at what is inside, what it does in the body, and what the evidence shows.
Chapter 5
Whether to use them, and which
Start with a question that comes before which sugar substitute: could you simply skip this sweet drink?
From best to worst, the ranking is short:
Best: water, tea, black coffee. With no sweetness in your mouth, the sweet receptors on your tongue are not pressed again and again.Next: whole fruit. It has sugar too, but the sugar arrives together with fiber, polyphenols, and vitamins.Then: swapping a full-sugar drink for one with sweetener. This saves sugar; it does not buy you health.Worst: getting through each day on full-sugar drinks.
Sweeteners are a bridge, not a destination. Switching from full-sugar soda to diet soda brings a small gain at the level of cutting sugar's direct harm, but it does not replace the thing that makes you want another bottle tomorrow: the expectation of sweetness.
Direction matters more than speed. Head toward depending less on sweetness; there is no need to get to zero overnight.
From best to worst, the ranking is short:
Best: water, tea, black coffee. With no sweetness in your mouth, the sweet receptors on your tongue are not pressed again and again.Next: whole fruit. It has sugar too, but the sugar arrives together with fiber, polyphenols, and vitamins.Then: swapping a full-sugar drink for one with sweetener. This saves sugar; it does not buy you health.Worst: getting through each day on full-sugar drinks.
Sweeteners are a bridge, not a destination. Switching from full-sugar soda to diet soda brings a small gain at the level of cutting sugar's direct harm, but it does not replace the thing that makes you want another bottle tomorrow: the expectation of sweetness.
Direction matters more than speed. Head toward depending less on sweetness; there is no need to get to zero overnight.
In practice · If you use one, which one
Which to use and which to skip: a practical order of decisionsQuestion 1: can you go without sweet drinks?
Yes (best): water, unsweetened tea, black coffee, sparkling water, homemade lemon water.No (realistic): use sweeteners to cut the harm of full-sugar soda; the diet version beats the regular one.This is the first decision: ask whether you need sweetness before asking which sweetener.
Question 2: if you use one, which?
Relatively preferred (weighing the evidence and your own response):
Whole fruit, and a very small amount of real sugar in home cooking; do not add sugar to drinks.Monk fruit (sweetened mainly by mogroside V): few studies, and so far no clear warning signal. That means little research, not proven safe. Good for desserts made at home.Stevia (high-purity rebaudioside A): a fair amount of short-term safety evidence, sparse long-term data; some people notice a bitter aftertaste.Allulose: so far it does not appear to raise blood glucose, but long-term data are sparse, and it is expensive and hard to find.
Use with conditions (fine short-term, cut back long-term):
Aspartame (diet sodas): the ADI is far above what most people actually consume; absolutely off-limits for people with PKU.Sucralose: the same; the newer question mark is sucralose-6-acetate (a manufacturing impurity that gut bacteria can also make from sucralose; Schiffman 2023's in-vitro screening reported it as genotoxic). That is a cell experiment, not human data.Acesulfame-K: usually blended with other sweeteners; no clear problem on its own.
Use with caution (new evidence has appeared):
Erythritol: the Witkowski 2023 signal has not been ruled out, so people with heart disease or a high risk of clots should cut back.Xylitol: Witkowski 2024 points the same way; if you have a dog, never let it get at xylitol.Other sugar alcohols (sorbitol, maltitol): they easily cause diarrhea.
Best avoided:
Products labeled 0 sucrose whose ingredient list still includes high-fructose corn syrup, maltodextrin, or fruit-juice concentrate: most are ultra-processed foods, not a healthy substitute.Children regularly drinking any "0 sugar" or diet drink: it shapes a preference for sweetness, and data in children are sparse.Large intakes during pregnancy: considered safe within the ADI, but given infant- signals like the one in Azad 2016, some caution makes sense.
In practice · How to cut back on sweetness
Question 3: why are you cutting back on sweetness?Goal one, fewer calories and less weight: sweeteners may help in the short term; long-term, the overall structure of your diet has to carry it.Goal two, blood-glucose control (diabetes, prediabetes): sweeteners are only one piece, to be weighed together with whole foods, exercise, and medication.Goal three, fewer cavities: xylitol gum and toothpaste are often used to prevent decay; there is some evidence, but it is not strong. If you have a dog, keep them out of reach.Goal four, less ultra-processed food: do not trade ultra-processed food made with sugar for ultra-processed food made with sweetener; the key is switching to whole foods.
Question 4: can you lower your threshold for sweetness?
The common advice: spend a few weeks (say, 4 weeks) cutting all sweet input at once, sugar and sweeteners alike.The first few days are usually the hardest, when the urge for something sweet is strongest.Many people say that after a while fruit tastes sweeter and a little sweetness is enough. But this has barely been tested in people: research cannot yet say whether eating less sweet food really changes how much you like sweetness, or how long it would take. Treat it as a self-experiment worth trying, not a guaranteed result.
Why the idea makes sense
Your tongue and the reward pathway behind it work, in many respects, on relative amounts. After a long soak in high sweetness, the same strawberry lands flat on that ruler. Bring the overall input down, and the same strawberry may seem sweet enough again. What changed is not the strawberry but the ruler you compare it with.
If the idea holds, quitting sugar by willpower undersells the problem: what is really moving is calibration. It would also explain why the first few days are the hardest: the ruler has not moved yet, but the gap is already there. The word if has to stay attached, though, until a trial measures it properly.
Myth · The trap of trusting sugar-free
"I drink sugar-free, so I'm healthy": the psychology of that thoughtMoral licensing
A set of randomized experiments by Khan and Dhar in 2006: after first making a restrained choice, people became more likely to pick the indulgent option next — as in I already had the diet drink, so I can have a slice of cake. The authors' explanation is that the first choice lifts your self-image a notch, and that notch becomes a permit for the next choice.Diet research shows a similar kind of compensation: the Mattes 2009 review notes that after swapping sugar for sweetener, people usually make up part of the calories saved.
The health halo
A 0 sugar label automatically suggests healthy, and by that logic it is easy to eat a bit more without noticing.The result: the original aim of 0 sugar, lose weight is partly canceled by the extra you eat.
Outsourcing willpower
"I can drink the diet one because it's a safe substitute for sugar": it feels as if the problem is solved, so you stop thinking about the overall structure of your diet.But the root problem — too much ultra-processed and processed food — has not moved.
The rebound from stopping all at once
Dropping all sugar and sweeteners at once leaves some people with headaches, irritability, and strong cravings for sweets in the first few days.It is not as dramatic as sugar addiction kicking in; it looks more like habit and the reward circuit being briefly thrown off balance.What helps: cut back gradually rather than all at once, and replace sweet input with other things, such as whole fruit, dark chocolate, or spices like cinnamon.
The perfectionism trap
"I must quit all sugar", then one slip makes you feel I've failed, and you give up entirely.For health behavior, doing it most of the time and keeping it up for the long run is more useful than perfect but on and off.
A view you can actually sustain
Sweetness is a taste. It is not love, reward, or a way to manage feelings.If you use sweet food to comfort yourself, celebrate, or cope with stress, that is more a behavioral problem than a nutritional one, and changing your diet alone will not reach the core.The real fix is finding rewards and ways of regulating emotion outside sweetness: exercise, time with people, making things, being outdoors, reading.
The aim of this story is not to teach you how to pick the perfect sweetener. It is to help you see what place sweetness holds in your life, and then adjust within what you can manage.
References · 12
- World Health Organization. (2023). Use of non-sugar sweeteners: WHO guideline. Geneva: WHO. Recommendation: WHO suggests that non-sugar sweeteners not be used as a means of achieving weight control or reducing the risk of noncommunicable diseases (conditional recommendation). The underlying systematic review identified 283 unique studies (50 RCTs, 97 prospective cohorts, 47 case-control studies). It does not cover people with pre-existing diabetes (out of scope), low-calorie sugars and sugar alcohols (polyols), or NSS in medicines and personal-care products (guideline PDF, WHO IRIS, Wayback snapshot 26 July 2025). www.who.int/publications/i/item/9789240073616
- Schiffman, S. S., Scholl, E. H., Furey, T. S., & Nagle, H. T. (2023). Toxicological and pharmacokinetic properties of sucralose-6-acetate and its parent sucralose: in vitro screening assays. Journal of Toxicology and Environmental Health, Part B, 26(6), 307-341. In-vitro screening reported genotoxicity for sucralose-6-acetate, a compound that arises as a manufacturing impurity of sucralose and can also be formed when gut bacteria acetylate sucralose. The work is cell-based screening, not human exposure data. 10.1080/10937404.2023.2213903
- International Agency for Research on Cancer. (2023). IARC Monographs evaluation of the carcinogenicity of aspartame, methyleugenol, and isoeugenol. Lyon: IARC. www.iarc.who.int/news-events/aspartame-hazard-and-risk-assessment-results-released
- Toews, I., Lohner, S., Kullenberg de Gaudry, D., Sommer, H., & Meerpohl, J. J. (2019). Association between intake of non-sugar sweeteners and health outcomes: systematic review and meta-analyses of randomised and non-randomised controlled trials and observational studies. BMJ, 364, k4718. In adults, non-sugar sweetener intake produced a small reduction in body mass index (mean difference -0.6; 95% CI -1.19 to -0.01) from two studies in 174 participants. The authors rate the certainty of evidence across outcomes as low to very low, and report no evidence of benefit for most health outcomes studied. 10.1136/bmj.k4718
- Joint FAO/WHO Expert Committee on Food Additives. (2023). Aspartame: summary of safety evaluation. 97th JECFA meeting. JECFA reaffirmed the ADI of 0-40 mg/kg body weight: with a diet soft drink containing 200 or 300 mg aspartame, a 70 kg adult would need more than 9-14 cans a day to exceed it, assuming no other intake; IARC classified aspartame as possibly carcinogenic to humans (Group 2B), citing limited evidence (WHO news release, 14 July 2023). www.who.int/news/item/14-07-2023-aspartame-hazard-and-risk-assessment-results-released
- Suez, J., Korem, T., Zeevi, D., Zilberman-Schapira, G., Thaiss, C. A., Maza, O., et al. (2014). Artificial sweeteners induce glucose intolerance by altering the gut microbiota. Nature, 514(7521), 181-186. Mostly mouse experiments: commonly used NAS formulations induced glucose intolerance through altered gut microbiota; the effect was abolished by antibiotics and transferred to germ-free mice by faecal transplant; the abstract reports similar dysbiosis and glucose intolerance in healthy humans but gives no human numbers. Not a randomised trial (abstract, PMID 25231862). 10.1038/nature13793
- Suez, J., Cohen, Y., Valdés-Mas, R., Mor, U., Dori-Bachash, M., Federici, S., et al. (2022). Personalized microbiome-driven effects of non-nutritive sweeteners on human glucose tolerance. Cell, 185(18), 3307-3328.e19. Randomised controlled trial, 120 healthy adults: saccharin, sucralose, aspartame or stevia sachets for 2 weeks at doses below the ADI, vs glucose-vehicle or no-supplement controls. Each NNS distinctly altered the stool and oral microbiome and the plasma metabolome; saccharin and sucralose significantly impaired glycaemic responses. Germ-free mice given microbiomes from top and bottom human responders largely reproduced their donors' glycaemic responses. Two authors are scientific co-founders of DayTwo (abstract, PMID 35987213). 10.1016/j.cell.2022.07.016
- Witkowski, M., Nemet, I., Alamri, H., Wilcox, J., Gupta, N., Nimer, N., et al. (2023). The artificial sweetener erythritol and cardiovascular event risk. Nature Medicine, 29(3), 710-718. The record id says Hazen, but Witkowski is the first author (Hazen is senior author and reports patents and consulting related to cardiovascular diagnostics). Observational: discovery cohort n = 1,157; in the US (n = 2,149) and European (n = 833) validation cohorts, top vs bottom quartile of plasma erythritol had adjusted 3-year MACE HR 1.80 (1.18-2.77) and 2.21 (1.20-4.07). Erythritol enhanced platelet reactivity in vitro and thrombosis in mice (FeCl3 carotid injury after an injection of 25 mg/kg). Pilot intervention: 8 healthy volunteers drank 30 g erythritol in 300 mL; plasma levels rose about 1000-fold and stayed elevated for more than 2 days. Erythritol is also made endogenously via the pentose phosphate pathway, so cohort levels mix intake and production (abstract, PMID 36849732; full text, PMC10334259). 10.1038/s41591-023-02223-9
- Witkowski, M., Nemet, I., Li, X. S., Wilcox, J., Ferrell, M., Alamri, H., et al. (2024). Xylitol is prothrombotic and associated with cardiovascular risk. European Heart Journal, 45(27), 2439-2452. 10.1093/eurheartj/ehae244
- Khan, U., & Dhar, R. (2006). Licensing effect in consumer choice. Journal of Marketing Research, 43(2), 259-266. Across a series of randomised experiments, making a virtuous choice (or merely intending one) increased subsequent preference for an indulgent or luxury option. The authors attribute the effect to a boost in self-concept that licenses the later indulgence rather than to changes in guilt. 10.1509/jmkr.43.2.259
- Mattes, R. D., & Popkin, B. M. (2009). Nonnutritive sweetener consumption in humans: effects on appetite and food intake and their putative mechanisms. The American Journal of Clinical Nutrition, 89(1), 1-14. Review: only about 15% of the US population over 2 years old ingests NNS; adding NNS to non-energy products may heighten appetite, but this is not seen when NNS come with other energy; substituting NNS for nutritive sweetener generally elicits incomplete energy compensation; there is no evidence of long-term efficacy for weight management and no benefit without energy restriction (abstract, PMID 19056571). 10.3945/ajcn.2008.26792
- Azad, M. B., Sharma, A. K., de Souza, R. J., Dolinsky, V. W., Becker, A. B., Mandhane, P. J., et al. (2016). Association between artificially sweetened beverage consumption during pregnancy and infant body mass index. JAMA Pediatrics, 170(7), 662-670. CHILD birth cohort (Canada), 3,033 mother-infant dyads; 29.5% drank artificially sweetened beverages in pregnancy, 5.1% daily. Daily vs none: infant BMI z-score +0.20 (0.02-0.38) and overweight at 1 year aOR 2.19 (1.23-3.88); no comparable association for sugar-sweetened beverages. Observational (abstract, PMID 27159792). 10.1001/jamapediatrics.2016.0301