It's Not the Chemicals, It's the Calories

People are missing the point about ultra-processed foods.

The word “ultra-processed” is doing a lot of heavy lifting when we talk about “ultra-processed food”. It conjures up images of huge factories, machines with vats and tubes, mixing and chopping and amalgamating. It evokes the idea of chemicals, additives, preservatives, colors, flavors. Fundamentally, whatever definition of “ultra-processed” you want to use, it suggests that the inputs to this mechanical behemoth are fundamentally different from the outputs. Food goes in… and something else comes out.

And ultra-processed food is undoubtedly bad for us. At least, we know that people who eat more ultra-processed food have worse health outcomes. And that feels right — it makes sense. Because ultra-processed food is unnatural, right?

When RFK Jr. was testifying before congress last week, he touched on the health impacts of ultra-processed food but buried the lede by focusing on those chemicals and additives, those colors and flavors.

He noted that our foods have more ingredients than similar products in Europe, that our FDA has a longer list of approved additives than similar agencies in other countries. The implication is that maybe the products would be ok, if we could just get rid of red dye number 40 or something.

This is not the problem with ultra-processed foods. Or, I should say, it is an extremely minor problem.

The main reason ultra-processed foods lead to poor health outcomes, is because we are very bad at not eating them.

These food products are designed, by very intelligent scientists, to be easy to eat, delicious, and shelf stable. We need to worry less about what effect red dye number 40 has on rats in a cage, and more on the effect it has on the visual appeal of the food product and how that overrides our usual ability to restrain ourselves.

The additives we need to worry about are the ones that make the stuff taste so good we can’t stop eating it. Sugar. Salt. Fat.

It’s really that simple.

In a game-changing 2019 study, 20 volunteers spent a month at the NIH. For two weeks, they ate a regular diet, then switched to one packed with ultra-processed foods. They could eat as much or as little as they liked, and both diets were matched for calories, macronutrients, and energy density. Still, on average, they ate an extra 500 calories a day on the ultra-processed diet. Simply put, these foods are bad for us because we just can’t stop eating them.

Maybe you think that as knowledgeable adults, we can learn to restrain ourselves. Or maybe you think that our habits have been too long engrained, and we’ll never really be able to say no to a bowl of Doritos. Either way, if there’s a group of people we really need to protect from ultra-processed foods, it’s our kids.

Multiple studies have demonstrated a link between ultra-processed food consumption and overweight and obesity in adults. The same can’t be said for kids — there simply isn’t as much data out there. Nevertheless, this is an incredibly important topic because, let’s face it, a lot of ultra-processed foods — breakfast cereals, fruit snacks, candy bars, yogurt drinks — are marketed directly at children.

Fortunately, this week we got a study from Kozeta Miliku and colleagues, appearing in JAMA Network Open, examining the effect of ultra-processed food consumption in very young kids, 3-year-olds, on rates of overweight and obesity at age 5.

Source: Chen et al. JAMA Network Open. 2025.

The study leverages data from the “CHILD” cohort study, a pregnancy-based cohort with 4 sites in Canada. At the 3-year visits, parents completed a food frequency questionnaire asking them how often their child ate any of 112 different individual items. Each item was assigned a processing level using the established NOVA system from unprocessed (like a tomato), to ultra-processed (like marinara sauce with added sugar and preservatives).

The question, of course, was whether consumption of more ultra-processed foods would be associated with higher weight two years later.

But before I tell you the answer, let’s look at some of the startling baseline data.

First of all, this graph shows that these 3-year-olds are getting nearly half of their caloric intake from ultra-processed foods.

Source: Chen et al. JAMA Network Open. 2025.

I was shocked by this at first, and then sobered, and then anxiously reflective as I tried to remember what kinds of things I was feeding my kids when they were that age.

The truth is, getting half your calories from ultra-processed foods is par for the course in North America. This adult study shows that, on average, we get 54% of our calories from ultra-processed foods. It was not always thus, I’ll point out, that same study shows a steady increase in calories from ultra-processed foods from 2003 to 2018.

But back to the kids. Among the more than 2000 children studied, a 10% increase in ultra-processed food intake was associated with a 0.07 increase in BMI, and a significant increase in waist-to-height ratio, subscapular skinfold thickness and triceps skinfold thickness.

They were also significantly more likely to be obese — an increase in obesity by about 20% for each 10% increase in ultra-processed foods.

This effect was really driven by the boys, the effects were much weaker among the girls.

But there’s a major issue here.

The authors report all their results adjusted for total energy intake. What they are showing us is the effect of ultra-processed foods on overweight and obesity if we subtract the fact that kids who eat more ultra-processed foods eat more total calories. This is crazy to me, and would lead, I think, to a vast underestimation of the effects. As I mentioned above, the real reason ultra-processed foods are so bad for us is that we eat more of them than we should — they cause us to ignore our own satiety signals and keep consuming — leading to a higher caloric intake. As I teach my epidemiology students, “don’t adjust for shit on the causal pathway”.

My fear then, is that people will look at this study and say: “well, sure, there’s a bit more obesity but it’s not that bad. In fact, in girls, it doesn’t seem to have much effect at all.”

That would be an inappropriate conclusion here. The study misses the worst part of these types of foods.

Now, you could say — wait — if they accounted for total calorie intake than these effects prove that it’s those additives and stuff that are causing overweight — a direct effect of the chemicals as opposed to the fact that the chemicals facilitate overeating. Perhaps — but as I said above I think this is a fringe concern. And the truth is, it’s hard to get an accurate measure of caloric intake from a food frequency questionnaire; real food diaries work much better for this.

So there is more work to be done on children, ultra-processed foods, overeating, and overweight. In the next set of studies, let’s not ignore caloric intake. It is not a factor that complicates our ability to determine the harm from ultra-processed foods, it is the primary mechanism of that harm.

A version of this commentary first appeared on Medscape.com.