Foods With Artificial Sweetener: The WHO's 2023 Verdict

Artificial sweeteners show up anywhere a label promises less sugar: diet soda, sugar-free gum, light yogurt, protein bars. The World Health Organization reviewed the evidence on them and settled one argument in 2023. Using non-sugar sweeteners to control body weight does not reliably work, and the guideline recommends against it for that purpose.^[1]

Dangerous is a separate question, with a separate kind of evidence behind it. The weight question had enough randomized trials behind it to answer with real confidence. The question of whether artificial sweeteners raise your long-term risk of disease is still running almost entirely on observational data, the kind that can show a pattern but can’t prove it caused anything. Here is what each side of that divide rests on, and what it means for the diet soda in your fridge.

Which sweeteners count as artificial, and which don’t

Six sweeteners have gone through the FDA’s formal food additive approval process: aspartame, acesulfame potassium, sucralose, neotame, advantame and saccharin (approved under older rules specific to that compound).^[2] These are what most people mean by “artificial sweetener,” and they’re the ones you’ll find in diet soda, sugar-free gum, light yogurt and protein bars.

Stevia leaf extract and monk fruit extract are a different case. Both reached the US market through the FDA’s Generally Recognized As Safe (GRAS) notification pathway rather than the food additive process.^[2] That’s a real regulatory distinction, not a marketing one: a plant-derived sweetener that a company self-affirms as safe under GRAS never went through the same petition review as aspartame did.

How US regulators classify common sweeteners

Based on the FDA's two approval pathways for low- and no-calorie sweeteners.

That distinction mostly disappears once you get to the research. The WHO’s 2023 guideline defines “non-sugar sweeteners” broadly enough to cover both the synthetic ones and stevia, and reviews them together for their effect on body weight and disease risk.^[1] For the rest of this article, “artificial sweetener” means the six FDA-approved ones specifically, since that’s what the strongest human studies measured.

One thing about the labeling itself is worth knowing: every one of those six sweeteners has to appear by its actual name in the ingredient list. That puts them in sharper contrast with natural flavors, a single catch-all term that can legally hide dozens of different compounds. If you want to know exactly what’s sweetening your food, the sweetener aisle is the easy case.

The WHO’s 2023 verdict, and what sits under it

The systematic review that the WHO guideline drew on looked at both randomized trials and cohort studies of non-sugar sweeteners and body weight.^[3] Among adults and children who were actively trying to lose weight, it found no measurable difference in body weight between those using non-sugar sweeteners and those who weren’t.

Among adults not specifically dieting, a handful of small, short trials found a modest benefit on body mass index, but the review rated that evidence very low to low certainty.^[3] That’s a thin reed to build a weight-loss claim on, and it’s the main reason the WHO guideline landed where it did: don’t recommend non-sugar sweeteners as a weight-control tool.^[1]

-0.73 kg average weight change across 14 randomized trials swapping sugary drinks for artificially sweetened ones PMID 41531289

A separate 2026 evidence review looking specifically at beverages found something more encouraging in the short term: across 14 randomized trials running 3 to 76 weeks, replacing sugar-sweetened drinks with artificially sweetened ones produced an average weight change of -0.73 kg and a body fat reduction of -0.72%.^[4] That’s the randomized-trial side of the story, and it’s a real, if modest, effect.

The same review then looked at 20 systematic reviews of long-running cohort studies, and found the opposite pattern: both sugar-sweetened and artificially sweetened beverages were associated with higher long-term risk of type 2 diabetes, cardiovascular disease and mortality.^[4] Short randomized trials and long cohort studies are answering two different questions, and they don’t agree.

Why the cohort studies can’t prove cause and effect

The best-known of those cohort studies is NutriNet-Sante, a French study that has followed more than 100,000 adults since 2009 using repeated 24-hour dietary records.^[5]

102,865 French adults followed for a median of 7.8 years in the NutriNet-Sante cancer analysis PMID 35324894

• Cancer. Among 102,865 participants followed for a median of 7.8 years, people with higher total artificial sweetener intake had a higher risk of overall cancer (hazard ratio 1.13), with the strongest signal for aspartame specifically (hazard ratio 1.15).^[5]
• Cardiovascular disease. A separate NutriNet-Sante analysis of 103,388 participants found total artificial sweetener intake associated with higher cardiovascular disease risk overall (hazard ratio 1.09), and a stronger link to cerebrovascular disease (hazard ratio 1.18), with aspartame again standing out.^[6]

A larger 2026 meta-analysis pooling 37 studies and more than 3.6 million participants found the same general pattern: higher artificial sweetener intake associated with increased risk of type 2 diabetes, several cardiovascular outcomes and all-cause mortality.^[7]

The catch: every one of those numbers comes from an observational cohort, where researchers tracked what people already ate rather than assigning anyone to drink diet soda. NutriNet-Sante measured exposure through self-reported dietary records and used statistical models to adjust for age, body mass index, smoking and overall diet quality, but that design can only show an association.^[5] It can’t rule out that people who already carry more metabolic risk are also the ones reaching for diet soda in the first place.

The WHO guideline makes almost the same point about its own evidence base. It cautioned that the link it observed might be confounded by participants’ baseline characteristics and complicated patterns of sweetener use, a category that could include people who already have weight or metabolic problems choosing diet products more often.^[1]

What the controlled trials show instead

Short randomized trials testing individual sweeteners tell a calmer story about blood sugar, at least at the doses people are exposed to in everyday products.

A 4-week randomized controlled trial in 42 healthy women found no significant differences in fasting glucose, insulin or oral glucose tolerance test results after daily saccharin, sucralose, or aspartame plus acesulfame potassium, compared with plain water.^[8]

A more striking trial complicates that reassurance. Researchers gave 120 healthy adults saccharin, sucralose, aspartame or stevia for two weeks, at doses below the acceptable daily intake.^[9] As groups, saccharin and sucralose significantly impaired glycemic responses; aspartame and stevia didn’t. The effect wasn’t uniform within the saccharin and sucralose groups either. It tracked each person’s own gut microbiome, and the researchers could partly predict who would respond by transplanting a participant’s gut bacteria into mice and watching the mice’s blood sugar shift the same way.^[9]

That’s a genuinely surprising result: two sweeteners from the same regulatory category can behave differently in your gut depending on which bacteria you’re carrying, in a way a blanket safety approval was never built to detect.

The researchers themselves are careful about what the finding does and doesn’t mean. The trial ran for only two weeks in 120 adults, and its own authors say the clinical implications of that personalized response still need future assessment, meaning it has not been shown to translate into different long-term weight or diabetes outcomes.^[9] A short mechanistic signal in a small group tells you something about mechanism. It does not yet tell you anything about population-level weight or diabetes risk over years.

What the evidence doesn’t settle

Put plainly, here’s where the science stops rather than where a headline wants it to:

• No trial has tested the hard outcomes directly. A review of the evidence behind the WHO recommendations concludes that no randomized trial has run long enough, or enrolled enough people, to confirm or rule out whether non-sugar sweeteners cause cardiometabolic harm, and it recommends restricting intake until longer studies exist.^[10]
• The personalized glucose response hasn’t been scaled up. The same two-week, 120-person microbiome trial that found individual differences in blood sugar response never followed anyone long enough to see if that difference changes real-world weight or diabetes outcomes.^[9]
• Your acceptable daily intake hasn’t moved. The WHO’s 2023 guideline is explicit that it isn’t based on toxicological safety assessments and isn’t meant to update or replace the acceptable daily intake levels that the FAO/WHO Joint Expert Committee on Food Additives and other regulators already set for each sweetener.^[1] The dose considered safe for aspartame or sucralose today is the same dose considered safe before 2023.

What this means for your diet soda, yogurt or protein bar

The WHO’s 2023 guideline recommends against relying on non-sugar sweeteners to control body weight or cut your risk of chronic disease.^[1] That’s a caution built on observational associations with long-term disease risk, not a finding that drinking one diet soda at an approved dose has been proven to cause disease on its own.

Bottom line: if you’re drinking diet soda or eating a sugar-free protein bar occasionally, no human trial has shown that approved-dose exposure alone gives you a proven cancer or cardiovascular risk. If you’re using artificial sweeteners daily specifically to lose weight, the WHO’s own review found little reason to expect that strategy to work long term, while the strongest cohort studies keep finding a disease-risk signal that nobody has yet explained away.

If you’re avoiding sugar for a personal sensitivity rather than for weight control, note that you’re solving a different problem than the one this guideline evaluated. The WHO’s recommendation is about non-sugar sweeteners as a weight and disease-risk strategy, not about whether sugar itself is the thing your body reacts to.

The practical move either way is the same one that works for every additive on this site: check the ingredient list for the sweetener’s actual name instead of trusting a front-of-pack “sugar-free” claim, since not every dairy product that says less sugar is sweetened the same way. Scanning that panel, rather than the marketing copy above it, is the whole point of the EatMatrix app.

This article summarises publicly available regulatory assessments and peer-reviewed research about food ingredients. It is general information and is not intended to diagnose, treat, cure, or prevent any disease, and it is not personalised dietary or medical advice. Regulatory positions change; the EatMatrix app reflects the current assessments in our database. If you have an allergy, a medical condition, or are pregnant, talk to a qualified clinician or dietitian before changing what you eat.