Two researchers at the University of Texas at Austin did something obvious that nobody had done at scale. They photographed the baby and toddler aisles at 21 Austin grocery stores, then assessed the nutrition information on every product marketed for kids 6 to 36 months. That came to 2,783 products. Infant formula, beverages, and first-food purees were left out, so this is the snack and meal end of the aisle, not the whole store.
81% were ultra-processed under the Nova classification. 49% failed at least one nutrient standard in the World Health Organization's Nutrient and Promotion Profile Model, guidance published in 2022 by WHO's European office with the WHO Collaborating Centre at the University of Leeds.
Erin Hudson and Marissa Burgermaster presented the audit at NUTRITION 2026 in July. Caveat up front, because it is load-bearing: this is a conference abstract. A committee of experts selected it, but the American Society for Nutrition's own disclaimer says it has not been through journal peer review and should be treated as preliminary. Strong signal, not a closed case.
Now the part that actually changed how I shop.
The standard those products were failing is not a standard here. The WHO model is guidance, not United States law, and it was written for Europe. The baby and toddler aisle in an American grocery store has no nutrient composition gate behind it.
The aisle is a marketing category, not a regulatory one
I had assumed, without examining the assumption, that the section with cartoon animals and age labels was gated somehow. That someone had drawn a line and these products were on the right side of it.
Something is regulated, just not what I thought. In January 2025 the FDA finalized action levels for lead in processed food for children under two, 10 parts per billion for most fruits, vegetables, and mixtures. A limit on a contaminant. No ceiling on sugar or sodium.
Hudson said the quiet part directly: "Many parents may assume that products sold on the baby/toddler food aisles of their grocery stores have to meet certain nutrient standards. Our findings highlight that half of the toddler foods available in local grocery stores do not."
The specific failures were not exotic. 25% of products exceeded the sodium threshold, 17% sugar, 12% energy density, 11% fat. The categories that tripped them are the ones in every cart I have ever seen: fruit pouches on sugar, oat bars on sugar and calories, mac and cheese and turkey sticks on sodium and fat, cheese puffs on calories, peanut butter puffs on sodium, fat, and calories.
None of that is a scandal on its own. Adult food fails nutrient thresholds constantly and we all know it, because adult food does not come with an implied certification stamped on the front. Toddler food does. The packaging is doing regulatory work that no regulator is doing.
99.4 percent
There is a peer-reviewed companion, and it is the one I would send to a friend.
In 2024, Daisy Coyle, Elizabeth Dunford, and colleagues at The George Institute for Global Health published an evaluation in Nutrients of 651 infant and toddler food products from the top 10 US grocery chains, collected in spring 2023 at eight stores in Raleigh, North Carolina, and from two online retailers. 60% failed the WHO model's nutritional requirements.
Then they scored the packaging against the same model's promotional rules. Zero percent passed. Not a low number. Zero.
99.4% carried at least one claim the WHO framework prohibits on food for this age group, an average of 4.4 per package, with one product carrying eleven. The most common were "non-GM" (69.9%), "organic" (59.3%), and "no BPA" (37.2%).
Read those three again. Not one tells you anything about sugar, sodium, or how processed the thing is. They are all true. They are all beside the point. The WHO framework prohibits them because they move a parent's hand toward the shelf while communicating nothing about what is inside.
The same paper found pouches were 47.3% of products surveyed, up from 6% of category sales in 2010 to 60% in 2023. That is a shift in share, not a claim about volume. Fewer than 7% of those pouches met the total sugar recommendation. The format that took over the category is the one that is hardest to evaluate by looking at it.
Does any of this reach development?
Here I have to be careful, because the honest answer is "possibly, a little, and the evidence is observational."
In 2026, the British Journal of Nutrition published an analysis of the 2015 Pelotas Birth Cohort, run by the Federal University of Pelotas in southern Brazil. First author Glaucia Treichel Heller and colleagues, two of whom also hold University of Illinois Urbana-Champaign appointments, characterized what 4,275 children ate at age 2 and compared it against cognitive scores at ages 6 and 7, measured with the Wechsler Intelligence Scale for Children, fourth edition.
An unhealthy dietary pattern at 2 (packaged snacks, instant noodles, sweet biscuits, sweets and candies, soft drinks, sausages and processed meats) was negatively associated with IQ, P below 0.001. In analyses the authors themselves label exploratory, high adherence was associated with a 2.24-point lower IQ (95% CI -3.35 to -1.13) versus low adherence, and 4.78 points lower (95% CI -7.06 to -2.49) among children who also had early-life growth deficits. A healthy pattern showed no significant association, which the authors attribute partly to how little it varied: about 92% of the children ate four or more of the foods that defined it.
What makes the finding worth reading is the adjustment list: child's sex, maternal age, schooling, work, and depression, family structure, socioeconomic status, household size, older siblings, preschool attendance, home stimulation score, exclusive breastfeeding duration, and whether food was introduced before six months. Those covariates came from a literature review and a directed acyclic graph, not from picking what made a point.
Now the limits, which are real. This is an association, not a causal finding, and no biological mechanism was tested. It is one cohort, and the authors say comparison to other countries should be made with caution. Parental IQ was not measured; home stimulation and early education stood in as proxies. A two-point difference in a population average is not a prediction about any individual child, and nobody should read it as one.
I am not going to tell you a snack determines a brain. I will say the size and care of that cohort make it hard to keep filing early diet under "does not matter," which is where I had it.
What I actually changed
Nothing dramatic, and nothing that belongs in a doctor's office. Your pediatrician is the authority on your child's diet, growth, and anything that concerns you specifically. This is about a shelf.
I stopped reading the front of the package. The front is advertising, and the 99.4% figure measures how good that advertising is at saying nothing. I read the ingredient list and the sodium and sugar lines instead. Takes a few seconds, works in any aisle.
I stopped treating the age label as a filter. "For toddlers" is a market segment, not a screen anything passed.
And I stopped treating a pouch as the same object as the fruit pictured on it, because I cannot see or chew what is inside and neither can Kaiser.
Hudson's own suggestion was warning labels on products that exceed the thresholds, which she thinks might also push manufacturers to reformulate. That would be useful. It does not exist, and I am not going to wait for it.
Why a development company is writing about groceries
Prodigy is a development coach, not a nutrition service. But the failure mode here is the one we built the product around.
It is not that toddler food is bad. It is that a whole category was designed to look pre-vetted, and parents reasonably read the design as a verdict. Hudson's audit measured the gap between what packaging implies and what anyone actually checked. That gap shows up in more than one aisle: the app store, the toy shelf, the class schedule.
The only defense is knowing which claims are load-bearing and which are decoration. That is research work, it is boring, and almost nobody has time for it. Our job is to have already done it before a parent has to.