A Canadian cohort links toddlers' ultraprocessed intake to later behavior problems

A cohort study of more than two thousand Canadian children has found that toddlers who consumed more ultraprocessed food at age 3 tended to score worse on a standard measure of behavioral and emotional symptoms two years later, at age 5 [s1]. The paper, published online 3 March in JAMA Network Open, adds a developmental-outcomes angle to a research literature that has mostly focused on ultraprocessed food's links to obesity and metabolic disease.

What the study measured

The researchers drew on the CHILD Cohort Study, a prospective Canadian pregnancy cohort that has followed mother-child pairs with dietary and behavioral data collected between September 2011 and April 2018 [s1]. The analysis included 2,077 children with both a diet assessment at age 3 and behavioral follow-up at age 5 [s1].

Ultraprocessed food intake was assessed through dietary reporting at age 3. Behavior was measured using the Child Behavior Checklist (CBCL), a validated parent-report instrument scored on a 0–100 scale, with higher scores indicating more adverse behavioral and emotional symptoms [s1].

What it found

Children with higher ultraprocessed food intake at age 3 had higher — worse — CBCL scores at age 5, indicating more behavioral and emotional symptoms [s1]. The study also modeled a substitution scenario: replacing ultraprocessed foods with minimally processed foods was associated with better behavioral outcomes [s1], a common technique in nutrition epidemiology for translating an association between one food category and an outcome into a more concrete, hypothetical statement about what a dietary swap might look like.

What the study can and can't tell you

This is a cohort study built on parent-reported diet and parent-reported behavior, not a randomized trial, and the distinction matters more than usual here. Diet at age 3 doesn't happen in isolation — it correlates with a household's income, parental stress, screen time, sleep habits, and dozens of other factors that independently shape a child's behavior at age 5. The study's authors accounted for some of these in their statistical models, but no amount of statistical adjustment in an observational cohort can rule out the possibility that some unmeasured factor — say, parental mental health or household chaos — is driving both higher ultraprocessed food consumption and worse child behavior, rather than the food itself causing the behavior.

Ultraprocessed food is also a broad category — industrially formulated products, from packaged snacks to sugary drinks to reconstituted meat products, that share manufacturing characteristics rather than a single nutritional profile. A diet high in these foods often means a diet correspondingly lower in whole foods, fiber, and certain micronutrients, and it's not possible from this study design to separate the effect of what's present in ultraprocessed foods from the effect of what's displaced by eating more of them.

The parent-report nature of both the exposure and the outcome is a further limitation: the same parent who reports what a 3-year-old eats is also the one scoring that child's behavior two years later, which opens the door to reporting biases that an objective measure — a lab-based behavioral assessment, for instance — would avoid.

Why it's still worth attention

None of those caveats make the finding meaningless. A sample of over 2,000 children with prospective, longitudinal data collection is a reasonably solid basis for an observational signal, and it lands alongside a growing body of research — largely in adults — linking ultraprocessed food intake to cardiometabolic and, in some studies, mental health outcomes. A study in young children extends that question to a developmental window when behavioral and emotional regulation is still forming, which is scientifically interesting independent of whether the mechanism turns out to run through the food itself, through what the food displaces, or through the household circumstances that shape both diet and behavior.

What to watch

The authors frame the practical implication cautiously: public strategies that promote minimally processed foods over ultraprocessed ones in early childhood could plausibly support behavioral development, without asserting that the current data prove causation [s1]. Confirmation would require either a randomized dietary intervention — difficult to run in young children over a multi-year horizon — or converging evidence from other cohorts with different populations and different confounding structures. Until then, this is one longitudinal association, from one cohort, using parent-reported measures on both sides of the equation — informative, but a single data point rather than a settled finding.

Sources

• Ultraprocessed Food Consumption and Behavioral Outcomes in Canadian Children — JAMA Network Open, published online 3 March 2026