The Ultra-Processed Majority

More Than Half

In August 2025, the Centers for Disease Control and Prevention published an analysis of dietary data collected from Americans between 2021 and 2023. The finding was straightforward and worth sitting with for a moment: 55 percent of all calories consumed by Americans one year of age and older came from ultra-processed foods.[1] Among children and teenagers ages 1 to 18, that figure rose to nearly 62 percent. Among young adults ages 19 to 39, it was 54 percent. Even the lowest-consumption adult group, those 60 and older, derived more than half their daily calories from the same category.[2]

This is not a crisis hidden in obscure nutritional data. It is visible in the grocery store, in the school cafeteria, in the vending machine, in the drive-through, in the center aisles of every supermarket in the country. The majority of what Americans eat every day is not food in the traditional sense, ingredients grown, raised, or produced close to their natural state and prepared in a kitchen. It is, increasingly, something manufactured: a product assembled from commodity inputs, modified starches, refined oils, flavor compounds, preservatives, and emulsifiers, engineered to be shelf-stable, cost-effective, and appealing enough to eat without thinking too hard about what it is.

The health implications of this shift are the subject of an expanding body of epidemiological research, and the pattern that research reveals is consistent and troubling. This article examines what ultra-processed food actually is, what the science says about its effects on human health, and why the direct-to-consumer, whole-food model at the heart of the Neighborhood Farms USA network represents something more than a culinary preference. It represents a structural alternative to a food system that has, over the course of a few decades, quietly changed what most Americans are actually eating.

What Ultra-Processed Actually Means: The NOVA System

Nutritional science has historically evaluated food primarily by its nutrient composition, the ratio of proteins to fats to carbohydrates, the vitamin content, the glycemic index. The NOVA classification system, developed in 2009 by epidemiologist Carlos Monteiro and colleagues at the University of Sao Paulo, Brazil, proposed a different lens: instead of asking what is in the food, ask what was done to it.[3]

NOVA divides all foods into four groups based on the degree and purpose of industrial processing. Group 1, unprocessed or minimally processed foods, includes fresh fruits, vegetables, meats, fish, eggs, milk, legumes, and grains in their natural or near-natural state. These foods may be washed, dried, frozen, or pasteurized, but nothing has been added and nothing essential has been removed. Group 2 covers processed culinary ingredients: oils, butter, flour, sugar, salt, the raw materials of home cooking. Group 3 is processed foods: products that combine Group 1 foods with Group 2 ingredients, typically to extend shelf life. Cheese, canned fish, salted nuts, pickled vegetables. These foods are recognizable and often contain two to five ingredients.

Group 4, ultra-processed foods, is a different category entirely. These are industrial formulations made primarily from substances extracted or derived from food, combined with ingredients rarely or never used in home kitchens: high-fructose corn syrup, hydrogenated oils, modified starches, protein isolates, artificial flavors, color additives, emulsifiers, stabilizers, and a range of other additives that serve industrial rather than culinary purposes. The finished product often bears little structural resemblance to any whole food ingredient. It is engineered to be hyperpalatable, convenient, shelf-stable, and inexpensive, and it is these properties, rather than any specific nutrient, that Monteiro argued make the NOVA Group 4 category a meaningful predictor of health outcomes.[4]

Critics of the NOVA system have raised legitimate methodological questions. The classification does not account for all cultural dietary contexts, and some observers note that the category boundaries are not always precise, a traditionally produced sourdough and a mass-market white bread may be classified differently despite sharing ingredients. The framework does not consider food access or affordability, which significantly shape what people actually eat. These are fair critiques, and NOVA is best understood as an epidemiological tool rather than a complete nutritional framework.[5] What it does exceptionally well is capture a pattern that nutrient-focused analysis misses: the overall degree to which a food has been industrially transformed, and what that transformation appears to do to the people who eat it.

“The extent and purpose of food processing may be more influential in determining a food’s health impact than its actual nutrient content.” — Carlos Monteiro, University of Sao Paulo, 2009

What the Epidemiology Shows

The 2024 BMJ Umbrella Review: The Largest Evidence Summary to Date
In February 2024, the British Medical Journal published what is, to date, the most comprehensive summary of the research on ultra-processed foods and human health: an umbrella review of 45 distinct pooled meta-analyses encompassing data from nearly 10 million participants.[6] The study was conducted by researchers at Deakin University in Australia, with no funding from companies involved in ultra-processed food production.

The review found direct associations between ultra-processed food consumption and 32 of the 45 health parameters examined, spanning mortality, cancer, and mental, respiratory, cardiovascular, gastrointestinal, and metabolic health outcomes. The strongest associations, classified by the researchers as “convincing evidence”, linked higher ultra-processed food intake to a 50 percent increased risk of cardiovascular disease-related death, a 48 to 53 percent higher risk of anxiety and common mental disorders, and a 12 percent greater risk of type 2 diabetes per additional serving per day.[7] These are associations, not established causal pathways, and the researchers acknowledged heterogeneity among study methodologies and the difficulty of ruling out confounding in observational data. But the breadth and consistency of the associations across independent studies of millions of people represents a pattern that the research community has taken seriously.

A January 2025 multi-center prospective cohort study published in The Lancet, drawing on data from 428,728 participants across nine European countries, found that higher ultra-processed food consumption was positively associated with all-cause mortality and mortality from circulatory diseases, cerebrovascular disease, ischemic heart disease, digestive diseases, and Parkinson’s disease. Replacing ultra-processed and processed foods with unprocessed or minimally processed alternatives was associated with meaningfully lower mortality risk.[8]

The NIH Randomized Controlled Trial: The Causal Evidence

Epidemiological associations, however consistent, cannot establish causation. What makes the ultra-processed food literature unusual is that there is also experimental evidence. In 2019, researchers at the National Institutes of Health published the first randomized controlled trial directly testing the effect of an ultra-processed diet on energy intake.[9]

Twenty adults lived in the NIH Clinical Center for four weeks. For two weeks they were provided an ultra-processed diet; for the other two, a minimally processed diet. Both diets were matched for total calories, energy density, macronutrients, sugar, sodium, and fiber. Participants could eat as much or as little as they chose. The result was striking: when eating the ultra-processed diet, participants consumed an average of 508 extra calories per day compared to when eating the minimally processed diet, equivalent to an additional full meal, and they gained an average of 0.9 kilograms over two weeks. On the unprocessed diet, participants spontaneously reduced their caloric intake and lost weight.[10] Critically, participants rated both diets as equally pleasant, suggesting the overconsumption was not driven by preference for the processed foods. Something in the structure of ultra-processed food itself, its eating rate, its lower protein satiety signal, its fiber composition, or a combination of these factors, was driving excess intake below the level of conscious awareness.

The Type 2 Diabetes Connection
Three large prospective U.S. cohort studies, pooling data from the Nurses’ Health Study, Nurses’ Health Study II, and Health Professionals Follow-Up Study, found that the highest consumers of ultra-processed foods had a 56 percent higher risk of developing type 2 diabetes compared to the lowest consumers, after adjusting for established risk factors.[11] Each additional serving of ultra-processed food per day was associated with a 5 percent higher risk. These are among the largest and longest-running dietary cohort studies in the world, and the consistency of the diabetes signal across them has influenced how public health researchers and clinicians think about dietary risk.

What Happens Inside: Additives, Emulsifiers, and the Gut

The epidemiological associations are clear enough that the scientific community has increasingly turned to the question of mechanism: what, specifically, do ultra-processed foods do to the body, and through what pathways?

The honest answer is that the full mechanistic picture has not been established. Ultra-processed foods are not a single thing, they vary enormously in their ingredient profiles, and the health effects documented in epidemiological studies likely reflect multiple overlapping pathways rather than a single cause. What the research has begun to identify are several candidate mechanisms that are biologically plausible and supported by experimental evidence, though human clinical data remain in early stages for some.

Emulsifiers, additives that prevent the separation of oil and water in processed products, found in everything from salad dressings to packaged breads to ice cream, have drawn particular attention. Common emulsifiers include carrageenan, polysorbate-80, carboxymethylcellulose, and lecithin. Preclinical studies and emerging human research suggest that some emulsifiers can disrupt the composition of the gut microbiome, increase intestinal permeability, and trigger low-grade inflammation, all of which have been linked to metabolic disorders and chronic disease.[12] A 2025 review in Clinical Gastroenterology and Hepatology and a 2025 narrative review in Nutrients both examined the evidence for emulsifiers and inflammatory bowel disease, finding biological plausibility through microbiome disruption and immune activation, while acknowledging that human clinical trial data at real-world doses remain limited and that the translation from animal models to humans requires caution.[13]

What is more firmly established is the overall nutritional profile that characterizes ultra-processed foods as a class: high in refined carbohydrates, sodium, and added sugars; low in dietary fiber, whole-food phytonutrients, and the diverse plant compounds that support gut microbiome health. These profile characteristics have well-established mechanistic links to metabolic dysfunction, regardless of questions about specific additives. A diet that chronically displaces fiber-rich, whole-food vegetables, legumes, and grains with products stripped of those components will, over time, produce predictable downstream effects on insulin sensitivity, microbiome diversity, cardiovascular risk markers, and inflammatory load. The structure of what has been taken away is as significant as the question of what has been added.

A Note on Seed Oils: Where the Science Stands

No discussion of ultra-processed food would be complete without addressing seed oils, which have become a significant point of public health debate. Seed oils, including canola, soybean, corn, sunflower, and safflower oil, are present throughout the ultra-processed food supply and are among the most widely consumed fats in the American diet.

The concern most frequently raised is that seed oils are high in omega-6 linoleic acid, which critics argue promotes systemic inflammation by converting in the body to arachidonic acid, a precursor to inflammatory signaling molecules. The mainstream epidemiological and clinical research does not support this pathway as stated. Multiple systematic reviews and meta-analyses of randomized controlled trials have found that increasing dietary linoleic acid does not significantly raise circulating arachidonic acid levels, only about 0.2 percent of dietary linoleic acid is converted through that pathway in humans, and does not increase markers of inflammation.[14] A 2025 review by researchers at Penn State, published in Nutrition Today, concluded that randomized controlled trials consistently show linoleic acid has a negligible impact on arachidonic acid levels and does not increase inflammation.[15] Studies measuring circulating linoleic acid directly in blood (rather than relying on dietary recall) have found higher levels associated with lower cardiovascular risk.[16]

Where the picture becomes more nuanced is in the context of ultra-processing itself. Seed oils in ultra-processed foods are frequently hydrogenated, a process that alters their chemical structure and is associated with negative cardiovascular outcomes independently of the oil’s fatty acid composition.[17] They also appear at high concentrations in a food matrix stripped of the fiber, antioxidants, and phytonutrients that whole-food sources of fat provide alongside them. The concern about seed oils in the American diet is therefore most accurately understood as a concern about the ultra-processed food matrix in which they predominantly appear, rather than a concern about linoleic acid as an isolated nutrient. The scientific consensus, as of 2025, is that replacing saturated fats with polyunsaturated fats from seed oils is associated with improved cardiovascular outcomes, while also being clear that consuming those same oils as part of a heavily ultra-processed diet pattern is associated with the health risks documented above.
“The problem is not any single ingredient in isolation. The problem is the overall dietary pattern that ultra-processed foods have created, one that increasingly crowds out the whole foods that human physiology was built around.”

How We Got Here: The Industrial Food Transition

The American diet did not always look like this. A century ago, ultra-processed foods as currently defined did not exist. The transition from a diet centered on minimally processed whole foods, prepared at home or in community, to one dominated by industrial food products happened gradually and then all at once, driven by a convergence of economic, technological, and cultural forces.
Following World War II, advances in food science, refrigeration, and mass distribution made it possible to produce and ship shelf-stable, standardized food products at a scale and price point that home cooking could not match. The food industry grew dramatically through the 1960s and 1970s, consolidating into fewer and larger companies with the capital to invest in flavor science, packaging technology, and mass-market advertising. By the 1980s, the center of the American grocery store had been transformed. By the 1990s and 2000s, fast food had become the dominant dining form for millions of families. The proportion of dietary energy derived from ultra-processed products rose consistently across every decade of this period.[18]

The consequences unfolded in parallel. Obesity rates in the United States, which held relatively stable through the 1970s, began climbing steeply in the 1980s and have continued rising. Today, more than 40 percent of American adults are classified as obese.[19] Type 2 diabetes prevalence roughly tripled between 1980 and 2020. Heart disease remains the leading cause of death. No single dietary factor explains these epidemics in full, and researchers appropriately resist simple causal narratives about complex chronic conditions. But the temporal correlation between the rise of ultra-processed food as the dominant calorie source in the American diet and the rise of these conditions has driven a generation of research, and that research has produced a consistent signal.

The Neighborhood Farm as Structural Opt-Out

The conversation about ultra-processed food is often framed as a matter of individual choice: read labels, cook at home, choose whole foods over packaged products. This framing is not wrong, but it places the burden of a systemic problem on individual consumers and ignores the structural conditions that make ultra-processed food the default. When 55 percent of all American calories come from a single food category, that is not primarily a failure of individual discipline. It is a reflection of a food system designed, at every level of production, distribution, and marketing, to make ultra-processed products the easiest, cheapest, and most available option.

The farms in the Neighborhood Farms USA network represent a different structural arrangement. A community farm selling direct to its community, through a CSA subscription, a farmers market stall, a farm stand, a neighborhood food hub, is providing something the industrial food supply chain is not built to offer: whole, minimally processed food, at or near the point of production, with minimal handling between harvest and table. There are no emulsifiers in a CSA box of greens. There are no modified starches in a dozen eggs picked up at a farm stand. The processing that occurs between the field and the consumer is, in most cases, exactly what NOVA Group 1 describes: washing, sorting, cooling, packing. Nothing is extracted. Nothing is added.

This is not incidental to what neighborhood farms do. It is definitional. A farm that grows food for direct local sale has no need to engineer shelf stability, optimize for the rigors of a 1,500-mile supply chain, or formulate for the hyperpalatability that makes industrial products competitive at retail. It needs to grow food that is fresh, flavorful, and ready to eat, which is precisely the food that the research on ultra-processed diets suggests we most need more of. NFUSA farms consistently grow a diversity of vegetables, herbs, legumes, and fruits that provide the fiber, phytonutrients, and whole-food complexity that ultra-processed diets systematically remove. The heirloom varieties, the seasonal abundance, the fermented products, the fresh-picked greens that characterize community farm shares are not luxuries. They are, increasingly, what the epidemiology says the human body is built to run on.
Making this opt-out accessible, rather than available only to those with the time, resources, and proximity to seek it out, is part of what NFUSA’s network-building mission is about. Every farm in our network that sells directly to its community, every CSA that a family joins, every farm stand that becomes part of a neighborhood’s food landscape is a small piece of a structural alternative to a food system that has, over several decades, made ultra-processed food the default. The research tells us clearly what the consequences of that default have been. The neighborhood farm, at its most essential, offers a different answer.