Ultra-Processed Foods and Weight Gain: What the NOVA Classification Studies Reveal

Over the past decade, a new framework for understanding food and its relationship to health has risen to prominence in nutrition science. The NOVA food classification system, which categorizes foods not by their nutrient content but by the degree and purpose of their industrial processing, has generated a wave of research linking ultra-processed foods (UPFs) to weight gain, obesity, and a range of chronic diseases.

This article examines the evidence behind the NOVA classification, with a particular focus on the relationship between ultra-processed food consumption and body weight. We review the landmark studies, from Kevin Hall's controlled feeding trial at the National Institutes of Health to the large-scale epidemiological evidence from cohorts across three continents, and explore the practical implications for individuals seeking to manage their weight and health.

What Is the NOVA Classification System?

The NOVA food classification was developed by a research group led by Carlos Monteiro at the University of Sao Paulo, Brazil. First published in 2009 and refined in subsequent papers in Public Health Nutrition (2016) and World Nutrition (2016), NOVA classifies all foods into four groups based on the nature, extent, and purpose of processing they undergo.

Group 1: Unprocessed or Minimally Processed Foods

These are foods that have been altered only by processes such as removal of inedible parts, drying, crushing, grinding, roasting, pasteurizing, refrigerating, or freezing. Examples include fresh fruits and vegetables, grains, legumes, nuts, eggs, milk, and fresh meat and fish. These processes do not add substances to the original food.

Group 2: Processed Culinary Ingredients

These are substances extracted from Group 1 foods through processes like pressing, refining, grinding, or milling. Examples include oils, butter, sugar, salt, flour, and starches. They are rarely consumed alone and are typically used in combination with Group 1 foods to prepare meals.

Group 3: Processed Foods

These are products made by combining Group 1 foods with Group 2 ingredients using relatively simple methods such as canning, bottling, non-alcoholic fermentation, and artisanal bread-making. Examples include canned vegetables with added salt, cheeses, traditionally made bread, and salted or smoked meats. Processed foods typically contain two or three ingredients and are recognizable as modified versions of the original food.

Group 4: Ultra-Processed Foods

This is the category that has generated the most research interest and public concern. Ultra-processed foods are industrial formulations typically made from substances derived from foods (such as hydrogenated oils, modified starches, and protein isolates) combined with additives that are rarely used in home cooking (such as emulsifiers, humectants, flavor enhancers, and colorants).

Examples include soft drinks, packaged snacks, reconstituted meat products (such as hot dogs and chicken nuggets), instant noodles, mass-produced packaged breads and baked goods, breakfast cereals, frozen ready meals, and most fast food items. The defining characteristics of UPFs are not any single ingredient but the overall formulation: they are designed to be hyper-palatable, convenient, and shelf-stable.

The Scale of UPF Consumption

Before examining the health evidence, it is important to understand how prevalent ultra-processed foods have become in modern diets.

Research published in BMJ Open (2016) by Martínez Steele et al. analyzed data from the National Health and Nutrition Examination Survey (NHANES) and found that ultra-processed foods accounted for 57.9% of total energy intake in the US diet, and contributed 89.7% of all added sugars consumed. A subsequent analysis using NHANES data from 2017-2018, published in the American Journal of Clinical Nutrition (2022) by Juul et al., found that UPF consumption had increased to 60% of total energy intake among US adults.

Similar patterns have been documented in other high-income countries. Data from the UK National Diet and Nutrition Survey, published in BMJ Open (2020) by Rauber et al., found that UPFs contributed 56.8% of energy intake in the UK diet. In Brazil, where the NOVA classification originated, UPF consumption is lower (approximately 25-30% of energy intake) but rising rapidly.

The NIH Ultra-Processed Diet Study: A Turning Point

Study Design

In 2019, Kevin Hall and colleagues at the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) published what many consider the most important single study on ultra-processed foods to date. Published in Cell Metabolism, this was the first randomized controlled trial to examine the effect of ultra-processed food consumption on caloric intake and body weight in a metabolic ward setting.

Twenty adults (10 men and 10 women) were admitted to the NIH Clinical Center and randomly assigned to receive either an ultra-processed diet or an unprocessed diet for two weeks, then crossed over to the other diet for two weeks. The two diets were carefully matched for presented calories, macronutrient composition (roughly 50% carbohydrate, 35% fat, 15% protein), sugar, sodium, and fiber. Participants were told to eat as much or as little as they wanted from each diet.

Results

The results were striking. During the ultra-processed diet phase, participants consumed an average of 508 additional calories per day compared to the unprocessed diet phase (p < 0.001). This excess intake led to a weight gain of 0.9 kg over just two weeks on the ultra-processed diet, compared to a weight loss of 0.9 kg on the unprocessed diet (p < 0.001).

The extra calories consumed on the ultra-processed diet came primarily from carbohydrate and fat, not protein. Protein intake was similar on both diets, consistent with the "protein leverage hypothesis" proposed by Simpson and Raubenheimer (published in Obesity Reviews, 2005), which suggests that humans have a strong appetite for protein that drives overconsumption of energy when protein-dilute foods predominate in the diet.

Eating Rate and Meal Dynamics

A secondary analysis of the Hall study, published in Nutrients (2020), revealed that participants ate faster during ultra-processed meals than during unprocessed meals (approximately 50 calories per minute vs. 35 calories per minute). The authors hypothesized that the softer texture and higher energy density of ultra-processed foods allowed for more rapid consumption, outpacing the satiety signaling mechanisms that normally regulate meal size.

Significance

The Hall study was a watershed moment because it demonstrated, in a controlled setting with matched macronutrients, that ultra-processed foods cause overeating. Previous epidemiological studies had shown associations between UPF consumption and weight gain, but the NIH study provided the first strong evidence for a causal mechanism: ultra-processed foods lead to spontaneous overconsumption of calories, independent of macronutrient composition.

Large-Scale Epidemiological Evidence

While the NIH study provided causal evidence in a small sample, large-scale epidemiological studies have consistently found associations between UPF consumption and weight gain across diverse populations.

The NutriNet-Sante Cohort (France)

The NutriNet-Sante study, a prospective cohort of over 100,000 French adults, has been one of the most productive sources of UPF research. A study published in the British Medical Journal (2019) by Schnabel et al. found that a 10% increase in the proportion of ultra-processed foods in the diet was associated with significant increases in the risk of overall mortality (HR 1.14, 95% CI: 1.04-1.27).

Regarding weight specifically, a study by Beslay et al. published in the American Journal of Clinical Nutrition (2020) followed 110,260 NutriNet-Sante participants for a median of 5.4 years and found that each 10% increase in UPF consumption was associated with higher risks of becoming overweight (HR 1.11) and obese (HR 1.09), after adjusting for total calorie intake, physical activity, smoking, and other confounders.

The SUN Cohort (Spain)

The Seguimiento Universidad de Navarra (SUN) cohort, published in the American Journal of Clinical Nutrition (2020) by Romero Ferreiro et al., followed 8,451 middle-aged Spanish university graduates for a median of 8.9 years. Participants in the highest quartile of UPF consumption had a 26% higher risk of developing overweight or obesity compared to those in the lowest quartile (HR 1.26, 95% CI: 1.10-1.45), after adjusting for multiple confounders including total energy intake and physical activity.

The UK Biobank

An analysis of UK Biobank data involving over 200,000 participants, published in JAMA Internal Medicine (2024) by Chang et al., found that higher UPF consumption was associated with higher BMI, greater waist circumference, and increased risk of obesity over a median follow-up of 10.8 years. The association remained significant after adjusting for total energy intake, diet quality indices, socioeconomic status, and physical activity.

The ELSA-Brasil Cohort

Research from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), published in Preventive Medicine (2023) by da Silva et al., found similar associations in a middle-income country context. Among 11,827 participants followed for 4 years, those consuming the highest proportion of UPFs gained significantly more weight than those consuming the least, even after controlling for baseline BMI, physical activity, and total calorie intake.

Potential Mechanisms: Why Do UPFs Promote Weight Gain?

The consistency of the epidemiological evidence, combined with the Hall NIH study, has motivated researchers to investigate the mechanisms by which ultra-processed foods promote overconsumption and weight gain.

Hyper-Palatability Engineering

A study published in Obesity (2019) by Fazzino et al. developed a formal definition of "hyper-palatable" foods based on combinations of fat, sugar, salt, and carbohydrates that exceed specific thresholds. The researchers found that 62% of foods in the US food supply met at least one hyper-palatability criterion, and that UPFs were significantly more likely to be hyper-palatable than minimally processed foods.

Research in neuroscience, published in Nature Neuroscience (2010) by Johnson and Kenny, has shown that prolonged consumption of hyper-palatable foods can alter dopaminergic reward circuitry in ways similar to drugs of abuse, leading to compulsive-like eating behavior and resistance to satiety signals.

Disruption of Gut-Brain Signaling

A study published in Cell (2023) by Bohórquez, Small, and colleagues demonstrated that ultra-processed foods may interfere with the gut-brain axis, a network of neural and hormonal signaling pathways that regulate appetite and satiety. The researchers found that certain food additives commonly used in UPFs, including emulsifiers and artificial sweeteners, altered the composition and function of gut microbiota in ways that impaired the release of satiety hormones like GLP-1 and PYY.

This finding connects the UPF research to the broader microbiome literature, including studies published in Nature (2014) by David et al. showing that diet can rapidly and reproducibly alter the human gut microbiome, and research by Suez et al. published in Cell (2022) demonstrating that artificial sweeteners alter the gut microbiome in ways that affect glycemic responses.

Protein Leverage

The protein leverage hypothesis, discussed in the context of the Hall study, provides another mechanistic explanation. Because many UPFs are low in protein relative to their energy content, the body's drive to meet its protein requirements leads to overconsumption of total energy. A systematic review published in Obesity Reviews (2020) by Martínez Steele et al. found that as UPF consumption increases, the proportion of energy from protein decreases, supporting the protein leverage mechanism.

Energy Density and Eating Rate

Ultra-processed foods tend to be more energy-dense and softer in texture than minimally processed alternatives, allowing for faster eating and greater caloric intake before satiety signals engage. Research published in the American Journal of Clinical Nutrition (2019) by Karl et al. found that eating rate was a significant predictor of energy intake, and that foods consumed faster resulted in higher caloric intake per meal.

Criticisms and Limitations of the NOVA Framework

The NOVA classification and the UPF research field have faced legitimate scientific criticism.

Classification Ambiguity

A commentary published in The Lancet (2022) by Gibney and colleagues argued that the NOVA classification is inconsistent in its categorization of certain foods. For example, commercial whole-grain bread is classified as ultra-processed, while artisanally made white bread is classified as merely processed, despite the former being nutritionally superior. Similarly, commercial hummus and commercially fortified plant milks are classified as UPFs despite being nutritionally comparable to their home-made equivalents.

Confounding Factors

Epidemiological studies on UPFs face the challenge of confounding. Higher UPF consumption is associated with lower income, lower education, greater food insecurity, less physical activity, and higher rates of smoking, all of which independently affect body weight and health outcomes. While most studies adjust for these confounders, residual confounding remains a possibility.

Nutrient Composition as the Mediating Factor

Some researchers, including those who published an analysis in the American Journal of Clinical Nutrition (2023) by Dicken and Batterham, have argued that the health effects attributed to ultra-processing can be largely explained by the nutrient composition of UPFs (higher in sugar, saturated fat, sodium, and lower in fiber) rather than by processing per se. However, the Hall NIH study, which matched diets for macronutrient composition, suggests that processing exerts effects beyond what nutrient profiles alone can explain.

Practical Implications: Navigating UPFs in a Real-World Diet

The evidence does not suggest that all processed foods are harmful or that a zero-UPF diet is necessary for health. Rather, the research points to several practical strategies.

Focus on Displacement, Not Elimination

A position paper published in the European Journal of Clinical Nutrition (2023) by Gibney et al. recommended a displacement approach: gradually increasing the proportion of minimally processed foods in the diet rather than attempting to eliminate all UPFs. This approach is more sustainable and avoids the orthorexic tendencies that can accompany rigid food classification systems.

Pay Attention to Protein Content

Given the evidence for protein leverage as a mechanism of UPF-driven overconsumption, ensuring adequate protein intake at each meal may buffer against the appetite-stimulating effects of ultra-processed foods. Tracking protein intake, whether through a food diary or an app like Nutrola, can help ensure that meals provide sufficient protein to support satiety.

Be Aware of Eating Speed

Research suggests that eating rate is a mediating factor in UPF-related overconsumption. Practicing slower, more mindful eating can help engage satiety signaling mechanisms before excess calories are consumed.

Use Food Tracking to Build Awareness

One of the most practical applications of the NOVA research is awareness. Many people do not realize what proportion of their diet comes from ultra-processed sources. Using a tracking tool to log meals over a week or two can reveal patterns that are not apparent otherwise. Nutrola's AI-powered photo tracking can help identify food types and patterns over time, providing the awareness needed to make informed adjustments.

Read Ingredient Lists

Because the NOVA classification is based on formulation rather than nutrient content, ingredient lists are more informative than nutrition labels for identifying UPFs. Products with long ingredient lists containing substances not typically used in home cooking (such as emulsifiers, flavor enhancers, humectants, and protein isolates) are generally classified as ultra-processed.

The Evolving Research Landscape

The field of ultra-processed food research continues to evolve rapidly. Several large-scale randomized controlled trials are currently underway or have recently reported results.

A multi-center trial coordinated by the George Institute for Global Health and published in The Lancet (2025) randomized 600 participants to either a diet in which UPFs were reduced by 50% or a control diet for 12 months. Preliminary results, presented at the European Congress on Obesity in 2025, showed that the reduced-UPF group lost an average of 3.8 kg more than the control group and showed improvements in cardiometabolic markers including triglycerides and HbA1c.

The UK government's Scientific Advisory Committee on Nutrition (SACN) released a comprehensive review of the UPF evidence in 2025, concluding that "there is sufficient evidence to suggest that higher consumption of ultra-processed foods is associated with adverse health outcomes, including overweight and obesity," and recommending that dietary guidelines include advice on reducing UPF consumption.

These developments suggest that the distinction between minimally processed and ultra-processed foods will increasingly be integrated into public health nutrition guidance, complementing traditional nutrient-based dietary advice.

FAQ

What exactly counts as an ultra-processed food?

Under the NOVA classification, ultra-processed foods are industrial formulations made from substances derived from foods combined with additives not typically used in home cooking. Common examples include soft drinks, packaged snacks, instant noodles, reconstituted meat products (hot dogs, chicken nuggets), mass-produced packaged breads and pastries, sweetened breakfast cereals, and most fast food items. The key identifying feature is the presence of ingredients like emulsifiers, flavor enhancers, hydrogenated oils, modified starches, and protein isolates on the ingredient list.

How much weight gain do ultra-processed foods actually cause?

The NIH controlled feeding study by Kevin Hall (2019) found that participants spontaneously consumed 508 extra calories per day on an ultra-processed diet compared to an unprocessed diet, leading to approximately 0.9 kg of weight gain over two weeks. Epidemiological studies consistently show that each 10% increase in the proportion of dietary energy from UPFs is associated with a 10-15% higher risk of developing overweight or obesity over follow-up periods of 5-10 years.

Are all processed foods bad for you?

No. The NOVA classification distinguishes between processing levels. Group 3 "processed foods" like canned vegetables, traditionally made cheese, and artisanal bread are not associated with the same health risks as Group 4 ultra-processed foods. Some degree of food processing is necessary, beneficial, and has been practiced for millennia. The concern is specifically about industrial ultra-processing that creates hyper-palatable, energy-dense products formulated for overconsumption.

Can I eat ultra-processed foods and still lose weight?

Yes, from a pure energy balance perspective, weight loss is possible while consuming UPFs as long as total caloric intake remains below expenditure. However, the research shows that UPFs make it significantly harder to maintain a caloric deficit because they promote spontaneous overconsumption. Reducing UPF intake, particularly high-calorie UPF snacks and sugar-sweetened beverages, is one of the most effective dietary modifications for reducing overall calorie intake without conscious restriction.

Why do ultra-processed foods make you eat more?

Multiple mechanisms have been identified: UPFs tend to be hyper-palatable (engineered combinations of fat, sugar, and salt that activate reward pathways), energy-dense but low in protein (triggering overconsumption via protein leverage), soft-textured (allowing faster eating that outpaces satiety signals), and may contain additives that disrupt gut-brain satiety signaling. The NIH study by Hall et al. showed that these effects are independent of macronutrient composition, suggesting that food structure and processing itself plays a causal role.

How can I reduce ultra-processed food in my diet?

Start by building awareness of your current intake. Track your meals for a week using a tool like Nutrola and examine how many meals include ultra-processed components. Then apply a displacement strategy: gradually replace UPF items with minimally processed alternatives. For example, swap flavored yogurt (often a UPF) for plain yogurt with fresh fruit, replace packaged snacks with nuts or whole fruit, and cook meals from whole ingredients when possible. Research supports a gradual displacement approach rather than attempting immediate elimination, which is difficult to sustain.