Only 14% of parents chose the yogurt with no added sugar or sweetener. With two front-of-pack labels, 60% did

Show American parents four yogurts with nothing extra on the front of the pack, ask them to pick one for their child, and 14% go for the one with no added sugars and no non-sugar sweeteners. Add a front label flagging added sugar and it's 29%. Add a second label flagging sweeteners and it's 60%. That's the headline from a randomised online experiment published ahead of print in the American Journal of Preventive Medicine on 22 October, by researchers at George Washington University and the universities of North Carolina and Washington.

The setup was simple. In November 2024, 2,896 parents of children aged 2 to 12 took an online survey and were randomly split into three groups: no front labels (990 parents), an added sugar label (949), or labels for both added sugars and sweeteners (957). Each parent saw four fruit drinks and four yogurts in random order. In each set, one product had neither added sugars nor sweeteners, one had added sugars, one had sweeteners, and one had both.

Look at the two steps separately and the interesting part appears. For yogurt, the sugar label moved the plain choice up 15 points, and adding the sweetener label moved it another 31. For fruit drinks the steps were 7 points and then 19. The design suggests one reason, though this is our reading and not a finding we can check. With a sugar label alone, the product sweetened only with non-sugar sweeteners carries no label, just like the plain one. To a parent scanning the front of the pack, the two look equally clean. The abstract doesn't say where the parents who left the sugary options went, so we can't tell how many moved to the sweetened product.

The labels changed what parents believed as well as what they picked. Asked which product was healthiest, 33% of parents in the no-label group named the yogurt without added sugars or sweeteners, 49% with the sugar label and 76% with both. Fruit drinks went 37%, 46% and 70%. Those figures always ran ahead of the choices. Across the three yogurt groups, 16 to 20 points more parents called that yogurt healthiest than chose it. These are group totals, so they don't track individuals, but they fit something most parents already know: the one you think is healthiest isn't always the one you put in the basket.

Now the limits, and they're big ones. This was a screen, not a shop. Nobody spent money, no child had a say, and nobody ate anything. The choice was also tidy in a way a real chiller isn't, with four options set up as a perfect comparison and the labels in plain view of people who knew they were in a survey. So 60% is what the labels did in that setting. It isn't a forecast of supermarket sales, and it says nothing about children's diets or health, which the study didn't measure. Parents who saw the sweetener label also rated sweetener-containing products as less healthy and said they were less likely to buy them, but those are intentions rather than purchases.

The most telling detail may be a change of words. When this work appeared as a preprint in May 2025, its summary called the winning products unsweetened and said the labels may support healthier choices. The published version calls them products without added sugars or sweeteners, and says potentially healthier. The percentages we compared are the same in both. That's a sensible edit, since a drink with nothing added can still contain sugar from its own ingredients, and we don't know from the abstract how these products compared on other nutrients.

It also matters because a warning's pull isn't a measure of the risk behind it. The World Health Organization's 2023 guideline recommends against using non-sugar sweeteners for weight control or to reduce disease risk, but it rates that advice as conditional, partly because the links to disease might be explained by who uses them and how. It doesn't apply to people with diabetes. So this experiment shows that a label can steer parents away from sweeteners. Whether that's a good swap for a particular child is a separate question, and one the experiment can't answer. A child with diabetes, in particular, shouldn't have their food or drink changed on the strength of it without their care team.

Could these labels reach real packs? In the US, not as things stand. The FDA's front-of-package box, proposed in January 2025, would rate saturated fat, sodium and added sugars as low, medium or high, and we found no final rule when we checked for this piece. FDA's description of it doesn't mention sweeteners, so a pack sweetened only with them could pass under the sugar line, which is the gap this experiment tested. Mexico already requires a warning reading contains sweeteners, not recommended for children. Until something similar exists elsewhere, the back of the pack is where to look. Sweeteners such as sucralose, aspartame or stevia appear in the ingredient list, even when the front says nothing about them.