Is Big Food Making Society Dumber? What the Evidence Actually Shows

Start with the fact that settles the underlying question, because it is not in dispute: adding iodine to table salt in 1924 raised measured IQ by roughly 15 points — a full standard deviation — among Americans living in the country's iodine-poor regions. Economists James Feyrer, Dimitra Politi and David Weil established this by comparing military test data from WWI and WWII conscripts, born before and after iodization, in areas that were naturally iodine-rich versus iodine-deficient. About a quarter of the population lived in deficient areas, so the national effect was around 3.5 points — roughly a decade's worth of the twentieth century's steady IQ rise, from one mineral added to one commodity.

So the premise holds. Nourishment moves population-level cognition, and it moves it by amounts large enough to see in national data. That is the strongest possible foundation for the question, and it is the reason this article exists.

The harder question is whether the modern food supply is running that experiment in reverse. Here the evidence is real but weaker, and this article separates the two carefully — because the version of this argument circulating online is considerably more confident than the data supports, and the honest version is unsettling enough without embellishment.

This article is educational, not medical or nutrition advice. Sakred Health is not affiliated with or endorsed by any company named here. Every figure comes from a named peer-reviewed source or a government publication.

Key statistics at a glance

• Populations in iodine-deficient US areas gained a full standard deviation of IQ — about 15 points — following 1924 iodization; the nationwide effect was roughly 3.5 points — Feyrer, Politi & Weil, NBER w19233 / JEEA, 2017
• Norwegian military conscript scores peaked with the 1975 birth cohort and have fallen about 0.2 points per year since — Bratsberg & Rogeberg, PNAS, 2018
• That decline appears within families — later-born brothers score lower than earlier-born brothers — which rules out genetic selection and immigration as explanations and points to environment — Bratsberg & Rogeberg, PNAS, 2018
• Adults drawing 20% or more of daily calories from ultra-processed food showed 28% faster global cognitive decline and 25% faster executive-function decline across a median 8 years, in more than 10,000 Brazilian adults — Gomes Gonçalves et al., JAMA Neurology, December 2022
• Ultra-processed foods supply 53.0% of calories for US adults and 61.9% for youth — well above the 20% threshold in that study — CDC/NCHS Data Brief 536, 2025
• In a controlled NIH feeding trial, an ultra-processed diet matched for sugar, fat, fiber and sodium still drove 508 extra calories per day — Hall et al., Cell Metabolism, 2019

The iodine precedent: proof that this is possible

Goiter was so common in the Great Lakes, Appalachian and Pacific Northwest regions that they were known collectively as the "goiter belt." The cause was simple — soil in those regions is low in iodine, and iodine is required to make thyroid hormone, which is required for normal brain development. In 1924, following a successful Michigan trial, salt producers began adding iodine.

What Feyrer, Politi and Weil recognized was that this created something rare in social science: a natural experiment with a clean before-and-after and a built-in control group. Men from iodine-poor counties and iodine-rich counties both took standardized military intelligence tests, in large numbers, in both wars. Iodization split the cohorts.

The gap closed. Men born after iodization in previously deficient areas scored about a standard deviation higher than their predecessors — while scores in already-sufficient areas moved comparatively little.

That is the single cleanest demonstration that a nutrient deficiency can impose a large cognitive cost on an entire population, and that removing the deficiency removes the cost. Anyone who wants to argue that food and public intelligence are unrelated has to explain this result, and they cannot.

The part nobody expects: the scores are falling again

Throughout the twentieth century, IQ scores rose roughly 3 points per decade across the developed world — the Flynn effect, named for the researcher who documented it. It is one of the most robust findings in psychometrics.

It has stopped, and in several countries reversed.

Bratsberg and Rogeberg, publishing in PNAS in 2018, analyzed three decades of Norwegian birth cohorts using military conscription records — a near-complete population sample rather than a survey. Scores peaked with men born in 1975 and have declined by about 0.2 points per year since.

The methodological move that makes this study matter is the within-family analysis. Two obvious explanations for falling national scores are compositional: different people are having children, or immigration has changed who is being tested. Bratsberg and Rogeberg tested brothers against brothers inside the same families. The decline showed up there too — a man born in 1985 scored lower than his own older brother born in 1975.

That rules out genetics and demographic change. Whatever is causing the decline is environmental, and it is acting on children growing up in the same households as their higher-scoring older siblings.

To be explicit about that chart: the study reports a rate, not a table of per-cohort scores, so the line is arithmetic applied to the published rate across the birth cohorts the researchers observed. It is drawn to show the shape and the slope, not to assert five independently measured data points.

The bars point in opposite directions even though two are labeled as gains and one as a decline — read the labels, not the lengths. The Norwegian figure covers the birth cohorts the study actually observed, roughly through the early 1990s. We have not extrapolated it forward to the present, because the study does not.

Where food enters — and how strong that link really is

Now the honest part, and it is where most versions of this argument overreach.

Bratsberg and Rogeberg did not identify nutrition as the cause of the reversal. They established that the cause is environmental and left the specific mechanism open. The candidates researchers have proposed include changes in education systems, declining reading of long-form text, shifts in media consumption, and changes in the tests' cultural fit. Diet is on the list. It is not established as the answer, and any article telling you otherwise is inventing a citation.

What does exist is a direct association between ultra-processed food and cognitive decline in adults. In December 2022, Gomes Gonçalves and colleagues published an analysis of ELSA-Brasil — more than 10,000 Brazilian adults aged 35 and up, followed for a median of eight years with repeated cognitive testing. Participants in the highest ultra-processed intake groups, at 20% or more of daily calories, showed 28% faster decline in global cognition and 25% faster decline in executive function than those in the lowest.

Grade this correctly: that is an observational cohort study. It shows association, not causation. People who eat more ultra-processed food differ systematically in education, income, physical activity and healthcare access, and although the researchers adjusted for many such factors, adjustment is never complete. It is genuinely suggestive. It is not the iodine result, and it should not be quoted as though it were.

The displacement argument

Here is the mechanism that makes the whole thesis coherent, and it does not require any exotic biology.

Ultra-processed food now supplies 53% of American adult calories and nearly 62% of youth calories. The concern is not only what those calories contain. It is what they crowd out. A stomach filled by engineered products is a stomach not filled by the foods that carry iodine, iron, zinc, choline, B vitamins and omega-3 fatty acids — the nutrients with the best-established roles in neural development.

That is the displacement argument, and it is the strongest form of the case: Big Food may not need to poison anyone to depress cognition. It only needs to occupy the plate. Iodine deficiency did not damage American brains by adding a toxin. It damaged them through the absence of something necessary.

It is worth noting this cuts against the simplest version of the story too. Fortification is itself an industrial food-processing intervention — iodized salt, enriched flour, fortified milk. Some of the greatest nutritional victories in history came from processing food, not avoiding it. The problem is not that food is manufactured. It is what gets optimized during manufacture, and satiety-per-nutrient has not been the target.

Where the argument outruns the evidence

The full chain some people draw runs: poor nutrition lowers intelligence, lower intelligence weakens the grasp of cause and effect, weaker causal reasoning erodes the public's ability to anticipate consequences, and a public that cannot anticipate consequences cannot hold institutions accountable.

The first link has strong support. The rest is a hypothesis, and we are labeling it as one.

No study has traced that path end to end, and it would be extraordinarily difficult to design one. Civic capacity is shaped by education, media structure, institutional trust, economic precarity and political design — variables that dwarf diet in most models. Presenting this chain as a research finding would be exactly the kind of overclaiming this blog exists to avoid.

What the evidence does support is the first and most important link, and it is remarkable on its own: the cognitive capacity of an entire population is a function of how well it is fed, and that has already been demonstrated at national scale, twice — once upward, once downward.

What this looks like on a plate

None of this argues for anxiety about individual foods. It argues for paying attention to the ratio — how much of the day's eating is doing nutritional work versus occupying space.

That is the same logic behind the rest of this section. The added-sugar disclosure gap explains why the ratio is hard to see at a counter. The controlled feeding trials explain why engineered food gets eaten in larger amounts without anyone deciding to. And our food chart sorts common items on nutrient density rather than on marketing category.

Frequently asked questions

Does nutrition really affect IQ?

Yes, at population scale and by large amounts. The clearest case is iodine: after the United States began iodizing salt in 1924, measured IQ in previously iodine-deficient regions rose by approximately 15 points — a full standard deviation — with a nationwide effect of about 3.5 points.

Are IQ scores actually falling?

In several developed countries, yes. Norwegian military conscript scores peaked with the 1975 birth cohort and have fallen about 0.2 points per year since. Because the decline appears between brothers within the same families, it cannot be explained by genetics or immigration and must be environmental.

Is ultra-processed food proven to lower intelligence?

No. A large Brazilian cohort study found 28% faster cognitive decline in adults getting 20% or more of their calories from ultra-processed foods, but that is an observational association, not proof of cause. The controlled evidence on ultra-processed food concerns appetite and calorie intake, not cognition.

Did researchers blame the IQ decline on diet?

No. Bratsberg and Rogeberg concluded that the reversal is environmentally caused but did not identify the specific cause. Proposed explanations include changes in schooling, reading habits and media consumption alongside diet.

Is processed food always the problem?

No — and the iodine story is the counterexample. Iodized salt, enriched flour and fortified milk are all industrial processing interventions, and they produced some of the largest public-health gains on record. The issue is what modern formulation optimizes for, not processing itself.

Sources

• NBER — Feyrer, Politi & Weil, The Cognitive Effects of Micronutrient Deficiency: Evidence from Salt Iodization in the United States (w19233)
• PubMed — Feyrer, Politi & Weil, Journal of the European Economic Association (PMID 31853231)
• PNAS — Bratsberg & Rogeberg, Flynn effect and its reversal are both environmentally caused (2018)
• JAMA Neurology — Gomes Gonçalves et al., Association Between Consumption of Ultraprocessed Foods and Cognitive Decline (2022)
• Cell Metabolism — Hall et al., Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain (2019)
• CDC/NCHS — Data Brief 536, Ultra-processed Food Consumption in Youth and Adults (2025)

Related: Do Ultra-Processed Foods Make You Eat More? — the controlled trial that removed willpower from the equation.