Ultra-Processed Foods: Understanding NOVA Group 4 and How It Makes You Overeat Without Realizing It

It is six o’clock on a weekday evening. You have just left work and stopped at the convenience store at the mouth of the soi. You pick up a cup of instant noodles, a pack of sausages, a bag of snacks, and a cold bottle of soda. The total is under one hundred baht. It fills you up, it is convenient, and there are no dishes to wash. After forty, meals like this happen more often than we think, because time is short, we are tired, and, frankly, they taste good.

Foods like these are known in nutrition as ultra-processed foods, or UPF. What makes this category important is that these foods can cause people to consume about 500 kcal more per day without realizing it, even when sugar, fiber, and salt are controlled to match ordinary foods. This finding comes from a real controlled laboratory study, not just a headline.

This article introduces the NOVA system, which classifies foods by their level of processing. It separates what is supported by evidence strong enough to use in life planning from what remains an association that cannot yet be stated with certainty. Because you are the one choosing what goes into your basket every day, you deserve to make those choices based on real evidence.

NOVA Classification: a system for classifying foods by level of processing

NOVA is a system that groups foods by their level of processing. It was developed by Prof. Carlos Monteiro and colleagues at the University of São Paulo during 2009 to 2010, and was published in Cadernos de Saúde Pública in 2010. Today, FAO uses NOVA and cites it in its 2019 report.

The idea behind NOVA is to look at “how much the food has passed through the factory” and “whether it contains ingredients not found in a home kitchen,” such as high-fructose corn syrup, hydrogenated oils, emulsifier, and stabilizer, rather than looking mainly at protein, fat, or carbohydrate content.

Group	Name	Familiar examples
1	Unprocessed / Minimally processed	Vegetables, fresh fruit, fresh meat, eggs, milk, brown rice, beans
2	Processed culinary ingredients	Oil, butter, salt, sugar (ingredients used for cooking)
3	Processed foods	Freshly baked bread, cheese, canned fish, pickled vegetables (a combination of groups 1 and 2)
4	Ultra-Processed (UPF)	Soda, sausages, instant noodles, packaged snacks, ready-to-eat frozen meals

The defining feature of group 4 is the presence of industrial ingredients, more than simply counting how many ingredients there are. The saying that “there are 5 or more ingredients” is only a common pattern often seen in these foods, not the main criterion in the original documents.

Another point to understand clearly is that there is still no clear evidence that “WHO has officially endorsed NOVA.” WHO accepts and cites NOVA, but that is not the same as a formal endorsement.

UPF makes people eat ~500 kcal more per day: RCT-level evidence

The strongest experiment on this issue is the 2019 study by Hall and colleagues, published in Cell Metabolism. It was conducted at NIH with 20 participants living in a controlled unit. Participants alternated between eating a UPF diet and an unprocessed diet, with calories, sugar, fiber, macronutrients, and sodium matched, and were then allowed to eat as much as they wanted.

The result was that during the UPF phase, people automatically ate 508 ± 106 kcal more per day. Weight increased by 0.9 ± 0.3 kg, and fat mass increased by 0.4 ± 0.1 kg. During the unprocessed-food phase, weight and fat decreased by similar amounts. Another RCT crossover study by Hamano and colleagues in 2024 found the same direction of effect.

Outcome	Confidence level	Evidence base
Excess intake of ~500 kcal/day	High	Multiple controlled RCTs
Increased weight and fat mass	High	Hall 2019 RCT
Insulin resistance	Moderate to high	meta-analysis and cohort studies (specific RCTs still pending)
Higher blood pressure	Moderate	observational cohort studies, no short RCT

The caution is that Hall 2019 was a short 2-week study with only 20 participants. We still do not know whether the effect persists over months or years. Evidence on insulin resistance and blood pressure comes from cohort studies, which found an increased risk of type 2 diabetes (RR = 1.74) and about a 23% increase in hypertension risk in the REGARDS study. But cohort studies cannot prove cause and effect 100%, because other confounding factors may be involved.

Some emulsifiers have detergent-like properties and can disturb the intestinal mucus layer

Polysorbate 80 (P80, code E433) and Carboxymethyl Cellulose (CMC, code E466) are emulsifier ingredients added to foods to create a smooth texture. Both have detergent-like properties, so they can disturb the mucus layer that keeps bacteria away from the intestinal wall.

Evidence from animal studies, including the 2015 Nature study by Chassaing and colleagues, found that P80 and CMC induced colitis, dysbiosis, and moved bacteria closer to the intestinal lining. The main mechanism came from changes in the microbial community, together with a reduced distance between bacteria and the intestinal wall.

Evidence in humans comes from a 2022 RCT in Gastroenterology that provided CMC at 15 grams per day for 11 days. It found microbiota encroachment into the inner mucus layer, where bacteria normally should not be present, along with reduced bacterial diversity and reduced beneficial compounds.

The caution is that not everyone responds the same way. Only some people are “responders” who are sensitive to these substances, and the amounts used in experiments are higher than typical real-world intake. The mechanism involves changes in the microbial community together with a reduction in the mucus layer. It does not always mean direct erosion.

The chain of mechanisms from leaky gut to chronic inflammation

Research lays out the chain of reasoning like this. A lack of fiber reduces beneficial bacteria such as Akkermansia muciniphila and F. prausnitzii. The body then loses short-chain fatty acids (SCFAs), such as butyrate and acetate, and the intestinal wall becomes weaker (leaky gut). This allows LPS to leak into the bloodstream, which then stimulates inflammatory markers TNF-α, IL-6, and CRP to rise.

There is real evidence supporting each link in this chain.

1. Fiber increases beneficial bacteria A systematic review of 29 RCTs found that inulin and FOS increased A. muciniphila and F. prausnitzii
2. SCFAs help reduce inflammation butyrate and acetate act as HDAC inhibitor, suppress inflammation through NF-κB, and help produce mucin
3. LPS stimulates inflammation LPS binds to the TLR4 receptor and activates NF-κB, producing TNF-α and IL-6

The caution is that a meta-analysis in patients with type 2 diabetes found that fiber supplementation did not significantly change IL-6 or TNF-α, even though LPS decreased. This may be because the studies were short and individual differences were high. The mechanism is therefore clear in animals, but human results are still mixed.

UPF and heart disease: multinational cohort evidence

UPF is linked to endothelial dysfunction, atherosclerosis, and increased cardiovascular disease (CVD) risk, with supporting evidence from multiple countries.

• Atherosclerosis The Aragon Workers’ Health Study in Spain (n = 1,876) found that people who ate 500 grams of UPF per day had 2 times the risk of subclinical coronary atherosclerosis compared with people who ate 100 grams per day (p < 0.001), after adjusting for heart risk factors
• CVD risk UK Biobank (n = 60,298 followed for 10.9 years) found that UPF increased CVD risk by 17%, increased coronary heart disease (CHD) risk by 16%, and increased all-cause mortality by 22%
• ACC 2026 Every 1 serving of UPF per day increased CVD risk by 5.1%, and eating more than 9 servings increased heart-related events by 67%
• Mechanism Fast-absorbed phosphate additive can cause vascular calcification through the hormones PTH and FGF-23, together with AGEs and emulsifier that damage the gut barrier, leading to body-wide inflammation

The evidence on this issue is supported by at least 4 independent sources, including Spain, England, Sweden, and consensus from ACC/AHA and ESC/EHJ, with no contradiction among them.

Important caveat: thyroid and autoimmune evidence is still only associative

This is the most important section, and it is why this article does not hide the uncertainty. You may have seen headlines claiming that UPF causes thyroid disease or autoimmune disease.

The truth is that the claim that UPF causes autoimmune hypothyroidism or demyelination through a leaky gut mechanism that leads to molecular mimicry has not been confirmed by a single human RCT. What we actually have right now is this:

• Thyroid An observational study from China (n = 8,732) found that UPF was linked to a 14% increase in subclinical hyperthyroidism per one SD, which is the opposite direction from what many people assume, hypothyroidism. The study on Hashimoto’s in Scientific Reports in 2020 was also only an association
• Demyelination A case-control study in Australia (Ausimmune, n = 775) found an 8% increase in first clinical demyelination per serving per day, and a Harvard cohort (BENEFIT, n = 451 patients with MS) found that the group with the highest UPF intake had a relapse rate about 30% higher than the lowest group. All of this is observational
• Molecular mimicry and food proteins The work by Kharrazian and colleagues in 2017 was in vitro only, and the authors themselves stated that it remains unproven whether consumption triggers inflammation
• Systematic review A 2024 Frontiers systematic review concluded that current evidence still cannot confirm a causal link between level of processing and adverse health outcomes

As for leaky gut in patients with autoimmune thyroid disease, the evidence is moderate. Microbial molecular mimicry, such as from Yersinia and H. pylori, has review-level evidence, but causality is still unclear. There is currently an ongoing RCT (NOVA trial, NCT06538831), but it is studying cardiometabolic outcomes, not thyroid outcomes.

How you can start

The evidence strong enough to plan around is that UPF really can make you overeat and gain weight, and it increases heart risk in studies from multiple countries. Some emulsifiers can disturb the gut in humans. Thyroid and autoimmune claims, however, are still only associations. You therefore do not need to panic over headlines that state more than the evidence can support.

3 steps you can take right away

1. Read the ingredient list on labels If you see names such as E433, E466, or names that do not exist in a home kitchen, that is a signal that the food is group 4
2. Move group 4 foods into the occasional-food category and make groups 1 and 2 the foundation of main meals, such as rice, eggs, fresh meat, vegetables, and fruit
3. Write down what you eat for 2 weeks Track your weight, bloating, and energy, and see how often convenience-store meals happen

For today, try just one meal first. The next time you stop by a convenience store in the evening, change from cup noodles and sausages to boiled eggs, plain milk, and one banana. You do not have to give up all processed foods in one day. Simply bring the foundation of your meals back to real food. Food decisions are personal, especially if you have thyroid disease or another medical condition. You should consult a doctor or dietitian, rather than deciding based on one article alone.