Aspartame: is it safe? A calm look at the 2023 evidence

Aspartame is one of the most studied food additives in the world, and in 2023 it produced a genuinely confusing news cycle: on the same day, two World Health Organization bodies published, one calling it a possible carcinogen, the other saying you could carry on drinking diet cola. Both were right, because they were answering different questions. This guide separates what is established from what remains contested, puts the numbers into real cans of drink, and flags the one group who must avoid aspartame entirely.

What aspartame actually is

Aspartame was discovered in 1965 and approved across Europe and North America through the 1980s and 1990s. Chemically it is the methyl ester of a dipeptide (a small two-part molecule) of two amino acids: L-aspartic acid and L-phenylalanine. Because it is around 200 times sweeter than sugar, it is used in milligram quantities where sucrose would need spoonfuls. On a label it appears as aspartame or its E number, E951, in diet soft drinks, sugar-free chewing gum, some tabletop sweeteners, certain yoghurts and a number of medicines.¹

Aspartame does not enter the bloodstream intact. It is fully broken down in the gut into its three parts before absorption: phenylalanine, aspartic acid and a small amount of methanol.⁴ This is the single most important fact for its safety, because none of the three is unique to aspartame. The amino acids are the same ones in any protein meal, in far larger amounts, and the methanol is the same compound your body handles from fruit: a glass of fruit or tomato juice typically delivers more methanol than a can of diet drink. The real question is therefore not whether these are "chemicals", but whether the small extra dose does anything ordinary food does not.

The 2023 headlines: what IARC and JECFA each said

On 14 July 2023, two WHO-linked bodies reported together.² The International Agency for Research on Cancer (IARC) ran a hazard assessment and placed aspartame in Group 2B, "possibly carcinogenic to humans". It cited limited evidence for cancer in humans (specifically hepatocellular carcinoma, a liver cancer), limited evidence in animals, and limited mechanistic evidence.³

The same day, the Joint FAO/WHO Expert Committee on Food Additives (JECFA) ran a risk assessment and reaffirmed the acceptable daily intake of 0 to 40 mg per kg of body weight, concluding there was no convincing reason to change it.² So one body said, in effect, "this could plausibly cause cancer and the human evidence is not strong enough to be sure"; the other said "at the amounts people consume, we do not see a problem". Both can be true at once.

Hazard versus risk: why Group 2B is not a danger rating

This is the crux. IARC classifies hazard: whether a substance is capable, in some circumstances and at some dose, of causing cancer. It deliberately does not weigh how much exposure it takes.² Group 2B is its second-lowest "possible" tier, a crowded category that also includes aloe vera whole-leaf extract and pickled vegetables. Group 1, the "sufficient evidence in humans" tier, contains tobacco, alcohol, processed meat and sunlight. The group tells you how strong the evidence is that something can cause cancer, not how dangerous it is in your fridge.

Risk is a different calculation: hazard combined with real exposure, meaning how much, how often, and in whom. That is the job of JECFA and, in Europe, EFSA, and on the doses people actually consume they judge aspartame acceptable.³ Sunlight is a Group 1 carcinogen and plain water can drown you, yet neither is a poison, because dose and context decide the danger.

The acceptable daily intake, counted in cans

EFSA's 2013 re-evaluation remains the most thorough European review. It examined the breakdown products in detail and confirmed the acceptable daily intake (ADI) at 40 mg per kg of body weight per day, judging it protective for the general population, including children and pregnant women, with PKU the one exception.⁴ JECFA uses the same 40; the US Food and Drug Administration uses a slightly higher 50. An ADI is a deliberately conservative line, set about 100 times below the highest no-effect dose in animal studies.¹

Put into practice, the numbers are reassuring. For a 70 kg adult, the ADI works out at about 2,800 mg a day. A 330 ml can of diet drink holds roughly 180 to 300 mg of aspartame, so reaching the ADI would take around 9 to 14 cans every day, assuming no aspartame from anything else.² EFSA's exposure surveys show even high consumers sit comfortably below the limit. That reframes the real question: not "what happens at the ceiling", which almost nobody reaches, but "is there harm at the modest amounts people actually drink". That is where the observational data come in.

The observational signals versus the trials

This is the genuinely unsettled area, and study type matters. The strongest recent signal comes from the French NutriNet-Sante cohort, an observational study, not an experiment: it watches what people already do rather than assigning them aspartame. Debras and colleagues, in PLOS Medicine in 2022, followed 102,865 adults and found higher aspartame intake associated with a small increase in overall cancer risk, with signals for breast cancer (hazard ratio 1.22, 95% confidence interval 1.01 to 1.48) and obesity-related cancers (hazard ratio 1.15, 95% confidence interval 1.01 to 1.32).⁵ A companion paper in the BMJ that year linked artificial sweeteners, including aspartame, to a somewhat higher rate of cardiovascular events.⁶

These are careful, large studies, but their design carries hard limits. Observational data show association, not causation, and are vulnerable to reverse causation (people worried about their weight tend to switch to diet products) and residual confounding (diet-drink consumers differ in many other ways). The effect sizes are also small: a 15 to 22% relative increase on a modest baseline risk. IARC's "limited evidence in humans" rested partly on a handful of such cohorts, with the clearest signal for liver cancer.³

Against that, the experimental side of the ledger, the randomised trials where aspartame is actually assigned and compared with a control, has not shown clear harm at normal intakes. Much of the older worry traces to lifetime rodent studies from the Ramazzini Institute in Italy, which reported increased lymphomas and leukaemias. Those studies were influential but heavily criticised on methodology, and regulators including EFSA judged them not reliable enough to change the ADI.⁴

Headaches, appetite and the gut

Headaches

Some people are convinced aspartame triggers their headaches, and this is the one symptom with a plausible signal in a minority. A small crossover trial by Van den Eeden and colleagues in 1994 tested 32 people who already believed aspartame gave them headaches, and found slightly more headache days on aspartame (33%) than placebo (24%).¹⁰ But it was small, had a high dropout rate, and the difference was driven heavily by a few participants. A later double-blind crossover study by Sathyapalan and colleagues in 2015 found no clear aspartame-specific effect.¹¹ So a genuine sensitivity in some individuals cannot be ruled out, but it is not established for the general population.

Appetite and weight

The idea that calorie-free sweetness confuses appetite is plausible and much discussed. Human trials are mixed, and systematic reviews generally find aspartame does not acutely raise blood glucose or insulin, nor reliably increase appetite. The WHO nonetheless advised in 2023 against using non-sugar sweeteners for weight control, because the long-term evidence does not show they help people lose weight and hints at possible downsides.⁷ The UK's Scientific Advisory Committee on Nutrition (SACN) reviewed the same guideline and was more cautious about how strong that evidence really is.⁸ The practical message: do not expect diet drinks to be a weight-loss tool, but the appetite-sabotage story is not proven either.

The gut microbiome

Animal work and some human studies suggest sweeteners might alter gut bacteria. For aspartame specifically, a randomised, double-blind crossover trial by Ahmad and colleagues in 2020 found no meaningful change in gut bacteria or short-chain fatty acids after two weeks of realistic intake in healthy adults.⁹ Other sweeteners such as saccharin have more suggestive microbiome data.

PKU: the one group who must avoid it

Here is the real, non-negotiable safety issue, and it is genetic rather than a matter of dose. People with phenylketonuria (PKU), an inherited disorder detected by the UK newborn heel-prick test, cannot properly process the amino acid phenylalanine. Because aspartame releases phenylalanine when digested, people with PKU must avoid it and manage all dietary phenylalanine carefully, since a build-up can cause brain damage.⁴ This is why UK and EU law requires any product containing aspartame to be labelled "contains a source of phenylalanine", alongside listing aspartame or E951.¹² For everyone without PKU, that warning is not relevant to them personally: the phenylalanine from aspartame is small next to a normal protein meal.

An honest bottom line

• Established: aspartame breaks down into ordinary food components; regulators worldwide judge it safe at the intakes people actually reach; and people with PKU must avoid it.²
• Reasonably settled: you would have to drink an implausible amount every day to approach the ADI, and aspartame does not acutely spike glucose or insulin.⁴
• Genuinely uncertain: whether habitual intake well below the ADI nudges long-term cancer or cardiovascular risk. Observational cohorts hint at small associations; this is not proof, and it is being actively studied.⁵
• Not supported: that aspartame at normal intakes is a proven cause of cancer, or that it is a useful weight-loss aid.
• Individual: a minority may get headaches; if you suspect that is you, a simple personal trial of avoiding it is reasonable.¹⁰

For most people, the calm reading is that an occasional diet drink is not something to fear; if you want to cut down, it is more useful to drink fewer ultra-processed drinks in general than to flee one molecule. Plain water, and not assuming "diet" means "healthy", does more than agonising over aspartame. For the metabolic and gut context, our health library covers blood sugar, weight and the microbiome, and our insights archive weighs individual sweeteners. New here? Our start here guide shows how to read any health claim calmly.

This article is educational and does not constitute medical advice, diagnosis, or a treatment recommendation. Medication uses described as “off-label” are not licensed for that purpose in the UK and should only be considered under qualified clinical supervision. Always speak to your GP, pharmacist, or a registered specialist before starting, stopping, or changing any treatment. If you have severe or alarm symptoms - unintentional weight loss, blood in your stool, difficulty swallowing, persistent vomiting, a fever, or severe pain - seek urgent medical care.