Artificial sweeteners: are they harmful, and do they help you lose weight?

Arguments about sweeteners are conducted as though the choice were between them and plain water. Usually the choice is between them and sugar, which is a different question with a different answer. Most of the confusion clears up once you know what the numbers on the safety side are actually measuring.

Two groups that keep getting confused

High-intensity sweeteners are hundreds of times sweeter than sugar, so they go into food in milligrams and contribute essentially no energy. In EU numbering this group covers aspartame (E951), sucralose (E955), acesulfame K (E950), saccharin (E954) and steviol glycosides (E960) — the last of these is what a label usually calls "stevia".

Sugar alcohols, or polyols — erythritol (E968), xylitol (E967), sorbitol (E420), maltitol (E965) — sit close to sugar in both sweetness and bulk. That is why they work in baking, where a few milligrams of sucralose cannot fill the space sugar used to occupy. They do carry some energy, and in quantity they pull water into the gut and are fermented there, which produces bloating and a laxative effect.

The distinction is worth holding onto, because nearly every argument about "sweeteners" is really about one group or the other. The safety debate — cancer, headlines, regulatory reviews — is about the intense ones. The complaint that a sugar-free chocolate bar ruined someone's afternoon is about the polyols, and it is not a safety question at all.

What an acceptable daily intake actually is

Every approved sweetener has an acceptable daily intake, and the number rarely appears in the arguments about it. The ADI is an amount per kilogram of body weight that a person can take in every day across a lifetime without expected harm. It is expressed per kilogram because a 45 kg adult and a 95 kg adult are not the same exposure at the same dose.

The derivation matters more than the number. Regulators do not start from a dose where harm was seen. They start from the highest dose at which no effect was observed at all in the toxicology studies — the no-observed-adverse-effect level — and then divide, conventionally by 100. That factor is two tenfold steps stacked: one for the possibility that humans are more sensitive than the test species, one for the possibility that some humans are more sensitive than others.

So the ADI is not a line between safe and dangerous. It sits roughly two orders of magnitude below the point where nothing measurable happened. Exceeding it on one day is not an event; the number was built for lifetime daily intake, with the margin already inside it.

How far ordinary intake sits below the limit

The arithmetic is worth doing once, because it settles the question better than any argument about it does. Take aspartame, which has the ADI most often quoted: 40 mg per kilogram of body weight per day. EFSA reached that figure in its 2013 re-evaluation, and the WHO/FAO expert committee on food additives has kept the same value.

Now convert the top row into drinks. EU rules cap aspartame in flavoured non-alcoholic beverages at 600 mg per litre, so a 330 ml can made at that legal ceiling holds just under 200 mg. Reaching the ADI means about fourteen such cans — not once, but as an everyday average sustained for years.

Real drinks are weaker than the ceiling. Typical soft drinks run in the range of 100 to 350 mg per litre, and many blend aspartame with acesulfame K rather than relying on it alone, which puts an ordinary can somewhere around 30 to 115 mg. On those figures the ADI is several dozen cans a day, every day. European intake surveys land where you would expect: exposure sits well below the limit even for the people who drink the most.

This is why the safety conversation and the practical conversation keep talking past each other. The regulatory margin on the intense sweeteners is enormous. Whether a sweetener is a good idea for a particular person is a question about habits and appetite, not about toxicology.

Aspartame, cancer research and the 2023 headlines

In July 2023 the WHO's cancer research agency placed aspartame in the category "possibly carcinogenic to humans". Headlines read this as confirmation of harm. The category means close to the opposite: the evidence for a link is limited, and the classification records that the question cannot be closed, not that an answer has been found.

That agency's job is hazard identification — can this substance cause the effect under any conditions at all. It does not assess dose, and its categories are not a ranking of danger. The same category holds things nobody organises their diet around, including aloe vera extract and traditional pickled vegetables.

The second half of the announcement was barely quoted. In the same week, the WHO/FAO expert committee on food additives reviewed the intake question and left the acceptable daily intake unchanged at 40 mg per kilogram. That committee answers the question people actually have: is the amount humans are getting dangerous. Two bodies, two questions, two answers that only look contradictory if you assume they were asked the same thing.

What the label is obliged to tell you

Under EU food information rules, a product sweetened this way carries "with sweetener" alongside its name, and two further warnings appear where they are needed.

The first is on aspartame. Where the ingredient is named, the label must read contains a source of phenylalanine; where it appears only as E951, the required wording is contains aspartame (a source of phenylalanine). Aspartame is built from two amino acids, and digestion releases one of them, phenylalanine. That line is addressed to people with phenylketonuria — an inherited disorder, screened for at birth across most of Europe, in which the enzyme that clears phenylalanine does not work. Left unrestricted, phenylalanine accumulates and damages the developing brain, so people with the condition follow a strict lifelong low-phenylalanine diet and must count aspartame into it. For everyone else the sentence carries no information whatsoever. It has the same shape as the gluten warning: a real and serious requirement for a small, identified group, routinely misread by everyone else as a general alarm — the mechanics of that misreading are set out in the piece on gluten.

The second warning is on the polyols. Where a food contains more than 10% added sugar alcohols, the label must state that excessive consumption may produce a laxative effect. This is not a formality. Sorbitol and maltitol in particular reach the large intestine largely undigested, and a "sugar-free" bag of sweets eaten in one sitting is a reliable way to find that out. People with irritable bowel syndrome tend to react at much smaller amounts, since polyols are one of the fermentable carbohydrate groups those diets restrict.

One more thing the label will not say plainly: sugar-free is not carbohydrate-free. Maltitol raises blood glucose appreciably, so anyone dosing insulin against a meal has to count it rather than assume a zero.

Do they actually help with weight loss?

The honest answer is more divided than either side of the argument wants. Swapping a sugar-sweetened drink for its zero-sugar version removes real energy from the day — a few hundred calories if the drink was not a one-off. For someone who drinks those daily, the swap is arithmetically sound and it works in short trials.

Against that, the WHO issued guidance in 2023 recommending against the use of non-sugar sweeteners as a means of weight control or of reducing the risk of non-communicable disease. Two qualifications belong with that sentence, because it is widely quoted without them. The recommendation is conditional, and the evidence behind it was graded low certainty — the WHO's own language for "this may change". And it applies to sweeteners used as a weight-management strategy: it does not extend to people with pre-existing diabetes, and it does not cover sugar alcohols.

The apparent contradiction resolves once you separate the drink from the habit. Trials that swap one beverage show a small weight reduction, because the calories genuinely leave. Long-term observational data show sweetener users carrying more weight, which is a good deal less surprising than it sounds — people typically switch to diet drinks because their weight is already a concern, and that reverses the arrow of causation. What survives both is the ordinary observation that the taste for sweetness and the habit of snacking are untouched by the swap, and over years those tend to decide the outcome. Weight change still tracks the energy deficit; a sweetener only ever removes one input to it.

Erythritol and a question that is genuinely open

In 2023 a study in Nature Medicine reported that people with more erythritol circulating in their blood went on to have more heart attacks and strokes. The association held across a discovery cohort and two validation cohorts in the United States and Europe, at roughly double the risk in the highest quarter against the lowest. The authors added laboratory work showing that erythritol at those concentrations makes platelets more prone to clotting, thrombosis experiments in mice, and a small pilot in eight volunteers whose blood erythritol stayed above the platelet threshold for more than two days after a single sweetened drink.

That is a more serious package than a lone correlation, and it still does not establish that eating erythritol caused the events. The body makes erythritol itself, from glucose through the pentose phosphate pathway, so a high blood level can be a marker of disturbed glucose metabolism rather than a record of what someone drank. Disturbed glucose metabolism is already a cardiovascular risk factor, which means that explanation has to be excluded before the first one can stand. The question is open, and saying so is the accurate position — not grounds for treating erythritol as established harm, and not grounds for waving the finding away.

A separate and firmer limit arrived the same year. EFSA re-evaluated erythritol and set a numerical acceptable daily intake for the first time: 0.5 g per kilogram of body weight per day, about 35 g for a 70 kg adult. It is derived not from cancer or cardiovascular data but from the dose at which erythritol causes diarrhoea, and unusually the figure was set level with that dose rather than a hundredfold below it. Two consequences follow, and they are what make erythritol the exception in this group. Because there is no safety factor, the number is a tolerance threshold rather than a distant ceiling. And EFSA found that 95th-percentile intake sits at or above it in every population group it assessed, not in one unlucky corner of the population — with the highest estimates among younger children. This is the one place in the group where the margin is genuinely narrow, and it is a limit on gut tolerance, not a toxicological one.

How much sugar counts as a lot

The reference point is the WHO's guidance on free sugars — those added during manufacture or cooking, plus the sugars naturally present in honey, syrups and fruit juice. The recommendation is to keep them below 10% of daily energy intake, with a further conditional suggestion to go below 5%.

At 2,000 kcal, 10% is 200 kcal, which is about 50 g of sugar: roughly one large bottle of a sugar-sweetened soft drink, and nothing else all day. The stricter target is about 25 g, which a single flavoured yoghurt and a coffee-shop drink can cover between them. Sugar inside whole fruit and in milk does not count toward either figure — it comes packaged with water, fibre or protein, and it is not what the guidance is about.

Sugar-sweetened drinks get singled out in this literature for a specific reason: liquid sugar arrives quickly, does not register as a meal, and is not compensated for later. That pattern is the one most consistently associated with fat accumulating around the organs and in the liver, which is the subject of the article on visceral fat.

What follows in practice

A sweetener is a transitional tool, not a destination. It earns its place where giving up a sweet drink is not happening on its own, and it does nothing at all where the sugar in a day is arriving through baked goods, breakfast cereal, sauces and dressings — none of which get reformulated by a tabletop sweetener.

The more useful move is to find out where the sugar in your day actually comes from before deciding what to do about it. It is almost never spread evenly. In most people's diets it concentrates in two or three sources, and at least one of them is somewhere they were not looking — a sauce, a "healthy" cereal, a daily coffee order. Once those are identified, whether to swap in a sweetener is a small decision rather than an argument about toxicology.

Count it from a photo

Frequently asked questions

Is aspartame harmful?

In 2023 the WHO cancer research agency classified it as "possibly carcinogenic to humans", a category that records limited evidence rather than an established link. In the same week the WHO/FAO expert committee on food additives reviewed intake and left the acceptable daily intake unchanged at 40 mg per kilogram of body weight. For a 70 kg adult that is about fourteen cans a day of the most heavily sweetened soft drink EU rules permit, sustained as a lifetime average.

Do sweeteners help you lose weight?

Replacing a sugar-sweetened drink with a zero-sugar version removes real calories, and short trials show a small weight reduction. The WHO nonetheless advises against using sweeteners as a weight-control strategy, on conditional guidance graded low certainty, because long-term benefit is not visible in the data. The taste for sweetness and the snacking habit are unchanged by the swap, and over years those usually decide the outcome.

Which sweetener is the safest?

The high-intensity sweeteners all carry an acceptable daily intake set roughly a hundredfold below the highest dose at which nothing was observed, and ordinary consumption does not come close to it. The practical differences are about tolerance rather than toxicology: sugar alcohols such as sorbitol and maltitol cause bloating and a laxative effect in quantity, which is why EU labels must warn about it above 10% added polyols. Erythritol is the one with a genuinely narrow margin, since EFSA set its limit of 0.5 g per kilogram of body weight from the dose that causes diarrhoea and found the heaviest-consuming children and adolescents above it.

Why do labels say aspartame contains a source of phenylalanine?

Aspartame is built from two amino acids and digestion releases one of them, phenylalanine. The warning is required so that people with phenylketonuria, an inherited disorder in which phenylalanine is not cleared and accumulates to the point of brain damage, can count it into a strict lifelong low-phenylalanine diet. It carries no information for anyone else.

How much sugar can I have in a day?

The WHO reference is free sugars below 10% of daily energy, with a conditional suggestion to go below 5%. At 2,000 kcal that is about 50 g, roughly one large bottle of sugary soft drink, or about 25 g for the stricter target. Free sugars include those added in manufacture and cooking plus the sugars in honey, syrups and fruit juice; sugar in whole fruit and in milk is not counted.

Read next

• Gluten in a healthy gut: what actually happens to it — Gluten has become the word people reach for to explain bloating, fatigue and stubborn weight, and also the word people say with a smirk.
• Alkaline water and “an acidic body”: what the stomach does to it — People talk about “an acidic body” as though it were a condition you can acquire from the wrong breakfast and correct with the right water.
• Calorie deficit: how to calculate it and how to hold it — A calorie deficit is the one thing without which weight does not come down.

This article is for general information. It is not medical advice, a diagnosis, or a prescription for treatment or a diet, and it does not replace a consultation with your doctor. If you have a health condition, are pregnant, take medication, or follow a diet prescribed to you, decisions about food belong with your doctor.

Figures from regulations, guidelines and studies are given as they stood when this article was prepared and may since have changed; check them against the primary sources. This article is not advertising, an offer, or individual advice, and neither the author nor the site owner is responsible for decisions taken on the basis of it.