Léo Ramos Chaves / Pesquisa FapespMonteiro, creator of the Nova food classification, on the steps of the School of Public Health at USPLéo Ramos Chaves / Pesquisa Fapesp
Carlos Augusto Monteiro, an epidemiologist from the School of Public Health at the University of São Paulo (USP), was one of the pioneers in documenting the nutritional transition in Brazil. He attributes the rapid growth of obesity worldwide to the increased food supply, especially industrialized foods designed to be cheap, strongly flavored, and almost always irresistible. In the innovative Nova classification proposed in 2009, which has helped change our understanding of how diet impacts health, he and his colleagues called these foods ultra-processed.
For over five years, Monteiro has been repeatedly named on lists of the most influential scientists in the world. In 2023, more of his scientific articles had been cited in the international academic literature than any other Brazilian, according to the AD Scientific Index. At the age of 76, he was awarded the title of professor emeritus by the School of Public Health and is formally retired from the university, although he continues to work at the Center for Epidemiological Research in Nutrition and Health (NUPENS), which he founded in 1990. He is currently leading NutriNet Brasil, a study that has been monitoring the diet and health of tens of thousands of Brazilians since 2020.
At the beginning of March, he spoke to Pesquisa FAPESP from his office on the second floor of the School of Public Health at USP about the global increase in obesity rates and one of the most recent studies into the impact ultra-processed foods have on health. Read the highlights from the interview below.
The world population has grown by 51% since 1990, but the number of children and adolescents with obesity has increased fourfold and the number of adults has more than doubled. Has the world completed the nutritional transition?
We are still in the midst of the nutritional transition—a change in the profile of the nutritional problems we face—with the simultaneous reduction in malnutrition and increase in the number of people who are overweight, obese, and suffer from chronic diet-related diseases. It’s analogous to the epidemiological transition, when infectious diseases gave way to chronic ones.
Will the situation get worse?
Probably. That’s the trend shown by the study published in The Lancet [see report]. A little while ago, we thought that the US had reached its ceiling, for example, with around a third of the adult population suffering from obesity. But even there, the problem continued to increase—now the prevalence of obesity has exceeded 40%. In the rest of the world it is lower, but the upward trend is occurring everywhere, which indicates that we may be heading towards a situation in which the disease will become the normal state. This doesn’t happen with other diseases. Being sick should be something unusual.
Disease should be an uncommon and unfortunate occurrence.
Exactly. It’s something unexpected, the result of a malfunction in the body. Natural selection protects species from the most frequent problems, but it is not perfect. A small proportion of people may have genes that make them more susceptible to certain diseases under certain conditions. But a small number of cases is very different from an epidemic, like we’re seeing with obesity, which usually results from something going very wrong in society.
In the case of obesity, what does this mean?
One of our body’s functions is to maintain a constant weight and a certain level of body fat. Both too much fat or not enough of it are harmful. For example, not having enough fat lowers a person’s ability to resist periods of food scarcity, while having too much can compromise their ability to escape danger. To keep fat levels stable, we need to ingest a similar amount of energy as we expend. All human beings have the ability to regulate this balance. But as with other biological functions, a small proportion of individuals are born with genetic characteristics that result in a less-than-optimal ability to regulate energy balance. This explains why there will always be a small proportion of people with obesity in any society. But it does not explain the dizzying growth of the disease.
The rise in obesity suggests we may be heading towards a situation where it will become the norm
There has been an important change in the environment.
Yes. A change that was not episodic, and that caused many people’s bodies to malfunction. It started in the US after World War II [1939–1945] and spread from there. Previously, only a small proportion of the population was obese: 2% or 3%. What changed? There is more or less a consensus that the main change, but not the only one, was in the food we eat. The supply of obesogenic foods [those which encourage weight gain] increased dramatically, and people began eating more calories than they needed. Not voluntarily—no one wants to be obese.
Why was it not possible to reach a middle ground, having overcome a situation in which hunger and malnutrition were common?
There has been a fantastic reduction in child malnutrition, especially in Brazil. As our group demonstrated, the change occurred here as a result of improvements in nutrition, basic sanitation, hygiene, vaccination, prevention of infectious diseases, and an increase in the population’s income. Malnutrition is partly linked to diet, but above all to poverty. In Brazil and other countries, the factors behind malnutrition have evolved positively. Now, people don’t become obese just because they now have more access to food. There has also been a major change in the food system.
What has changed?
Private companies—those that produce much of the food we consume—simply want to maximize profit. The industry buys raw materials from farmers, processes the food, adds some value to it, and extracts the profit. It pasteurizes milk, for example, which increases shelf life and makes it safer to consume, or transforms it into cheese, which makes it last longer and gives it different sensory properties that are more pleasing to the palate. So far, so good. Human societies have been processing food for millennia. The food industry also produces culinary ingredients, like sugar, oil, salt, and butter, which are essential for preparing food. But advances in food science and technology in the second half of the last century allowed the industry to exponentially increase its profitability. This is what gave rise to ultra-processed foods.
What exactly are ultra-processed foods?
They are not exactly food—they are more like formulations of chemical compounds. Some of these compounds are carbohydrates, proteins, oils, and fats extracted from high-productivity crops, such as soybeans, corn, wheat, and sugarcane. Others are carbohydrates, proteins, oils, and fats subjected to chemical changes, such as modified starches, hydrolyzed proteins, and hydrogenated oils. Others still are flavorings, colorings, emulsifiers, thickeners, stabilizers, and other essential additives used to make the recombination of the previous compounds taste more like whole foods or culinary preparations of those foods. Ultra-processing distorts a food’s characteristics. It treats them as a raw material that will provide the ingredients for the production of a new food. They are therefore a formulation of chemical substances, not a combination of foods.
What’s the difference?
Spaghetti al pomodoro is a combination of foods. It has pasta, tomatoes, basil, garlic, and grated cheese. Instant noodles, however, are a formulation containing starch, oils, and fats with the addition of flavorings, colorings, and stabilizers to resemble the traditional culinary preparation and even to surpass it in terms of sensory properties that satisfy the consumer. With an almost infinite palette of low-cost ingredients, the ultra-processed industry can produce irresistible alternatives for practically all foods and culinary preparations. It’s a gold mine.
The result is not the same as natural food?
Imagine that a component used in car production is made of steel or aluminum. If someone were to discover a type of plastic that has the same properties but costs less, the entire industry would adopt it. It would make cars lighter and more economical. They would be cheaper for the consumer and more profitable for the industry. But food is different from cars. Food is ingested, absorbed, metabolized, and incorporated into our organism, which evolved to eat real food that needs to be chewed, takes time to make you feel satiated, and has a certain energy density [concentration of calories]. Our bodies have a very good design that even allows us to deal with some changes, but there is a limit. Ultra-processed products go beyond our capability for adaptation. They try to imitate food and confuse our bodies. Properties such as taste, aroma, flavor, and color, which we used to identify foods and their nutritional composition, do not have the same meaning. There is a discrepancy between appearance and content. The food industry will always seek to maximize its profit, and in this sense, ultra-processing technology is unbeatable. The problem is that it can turn food, something essential to our health, into something that causes illness.
Ultra-processing decharacterizes food and treats it as a raw material to provide the ingredients for the production of a new food
This strategy was born in the US and later became universal. How did that happen?
The business model of producing and selling ultra-processed foods began in the 1980s with the growth of multinational corporations, enabled by the deregulation of domestic markets and globalization of the economy. In this landscape, multinational food industries can purchase ultra-processed ingredients from wherever they are cheapest, combine them wherever the tax rules are most advantageous, and sell the final product all over the world, with singular marketing campaigns. Globalization of the economy has helped advance this business model, in which corporations prevail over other economic agents.
How is this all linked to the obesity epidemic?
Because of this business model, there has been a massive change in the profile of food supply and the way it is advertised. The change was progressive. Until the late 1980s, the prevalence of obesity in the US adult population was about 15% and slowly increasing. Then it started growing much more quickly until it reached the current level of more than 40%. Did ultra-processed foods not exist before? They did, but not on the same scale. Food has changed so much that people’s’ ability to adapt may have been exhausted.
In March, Brazilian researchers published a review article in Nature Metabolism highlighting environmental factors that determine obesity. How could this help tackle the problem?
From an individual standpoint, obesity is caused by a combination of genetics and environment. A person’s genetic characteristics make them more or less likely to be obese. But the outcome is not set in stone. That depends on the individual’s environment. From a populational and societal standpoint, it takes a huge change in the environment to cause our organism, which is so well designed, to fail. As the authors of the article point out, several changes occurred simultaneously that stimulated the development of obesity.
How did your group arrive at the idea, presented in 2009, that the degree to which food is processed could contribute to weight gain?
In the early 2000s, we were studying the evolution of malnutrition over time, which had fallen dramatically in Brazil, and seeing an increase in people being overweight and obese. We began to wonder what might explain it. We started looking at what was changing in people’s diets and analyzed data from food purchase surveys carried out by the IBGE [Brazilian Institute of Geography and Statistics] since the mid-1970s. The first thing that caught our eye was the substantial drop in purchases of culinary ingredients such as salt, sugar, and oil, as well as the decline in purchases of rice, beans, and cassava. These reductions were offset by the increasing popularity of a seemingly heterogeneous set of products, including instant noodles, sausages, and soft drinks. All ready-to-eat products. It occurred to us that these changes could be reflecting the fact that traditional eating habits based on cooking with whole foods were being replaced by a desire to buy products that require little or no preparation. These ready-to-eat products were not foods modified by processing, but formulations of substances resulting from food processing. In 2009, in a comment published in the journal Public Health Nutrition, we called this group ultra-processed foods.
This is the term you use in the Nova classification, which separates foods into four groups: unprocessed or minimally processed foods, processed culinary ingredients, processed foods, and ultra-processed foods.
That’s right. In 2009, we presented what was the embryo of this classification. It included just three groups, with the third yet to distinguish between processed and ultra-processed foods. In 2012, the Brazilian Ministry of Health asked us to review the Brazilian Dietary Guide, published in 2006. In that project, we separated ready-to-eat foods that were modified whole foods, such as preserves, traditional breads, and cheeses, from formulations of food-derived substances, such as instant noodles, sausages, and soft drinks. And that was the birth of the Nova classification with its four food groups, which guided the recommendations of the 2014 Brazilian Dietary Guide.
What have your studies shown about the effect of ultra-processed foods on health?
Our first investigation, carried out using data from a national survey on food consumption in Brazil and published in Preventive Medicine in 2015, found that obesity was more common among people who had a diet high in ultra-processed foods than among those who consumed less of these types of food. In the following two years, two studies were published by other groups overseas, describing a cohort of around 10,000 Spaniards followed for 10 years, which demonstrated the prospective association between the consumption of ultra-processed foods and the risk of developing obesity and arterial hypertension. Between 2018 and 2020, researchers from the University of Paris-Sorbonne, working in collaboration with our group, published a series of longitudinal studies involving a cohort of more than 100,000 French people followed for around 10 years. These studies demonstrated a prospective association between the consumption of ultra-processed foods and the risk of developing obesity, diabetes, cardiovascular disease, breast cancer, and depression. Since then, dozens of similar studies carried out in the USA, UK, Italy, the Netherlands, South Korea, China, and Brazil have confirmed the link between ultra-processed foods and the risk of chronic diseases, including kidney, gastrointestinal, and lung diseases.
Studies have shown a link between the consumption of ultra-processed foods and the risk of developing obesity, diabetes, cardiovascular disease, breast cancer, and depression
The most recent article on ultra-processed foods was published in The BMJ in March. It analyzed several other studies and found an association between ultra-processed foods and 32 health problems, in particular increased mortality and problems related to the heart, metabolism, and mental health.
It was an umbrella review, which summarized the findings of several other reviews that used meta-analyses. In this case it was 45 meta-analyses. The meta-analysis umbrella review is the gold standard of observational studies (those in which the researchers do not intervene in the population studied). Something important to note is that this study classified the quality of the evidence demonstrated in each meta-analysis. Convincing or highly suggestive evidence of an association was demonstrated for diseases such as diabetes, obesity, and premature mortality of all causes. Less strong evidence was demonstrated for other health problems. None of the 45 meta-analyses showed that ultra-processed foods protect against disease.
How much do these studies contribute to our understanding of the impact ultra-processed foods have on health?
They are essential. It was through observational, prospective association studies that it was discovered that tobacco causes lung cancer. They make it possible to monitor a very large number of people, representative of the population, under normal living conditions. A single associative study is not enough to demonstrate that a certain factor causes a certain disease, but when multiple studies carried out in different countries by different researchers confirm the association, the probability is higher that the factor is a real cause of the disease. And the probability is even greater when other studies indicate plausible mechanisms to explain why exposure to a certain factor increases the risk of the disease. Several studies have suggested plausible mechanisms for the damage ultra-processed foods cause to health, including their unbalanced nutritional profile, high energy density, lower capacity to satiate the appetite, negative influence on the intestinal microbiome, and more. Ideally, the results of observational studies should be confirmed by clinical trials in which the participants are randomly assigned to groups and asked to consume high or low amounts of ultra-processed foods. Logistical and ethical problems, however, make clinical trials on nutrition and chronic diseases more complex. It is not really feasible to ask people to consume a certain diet for the length of time necessary to observe the onset of a chronic disease. It’s also unethical when it is known that the diet in question is potentially harmful to health.
To date, there has only been one clinical trial involving ultra-processed foods, run by Kevin Hall of the US National Institutes of Health.
That’s right. It involved 20 adult participants who had been admitted to a hospital. For two weeks, half of the participants received a diet containing 80% ultra-processed foods and the other half received a diet with no ultra-processed foods. In the two weeks after that, the groups were reversed. The experiment showed that while eating the ultra-processed diet, the participants consumed about 500 kilocalories more per day than when they were given the ultra-processed-free diet. That number of calories represents about a third of the daily total that an adult should eat. In the two weeks of eating a diet high in ultra-processed food, the participants gained around one kilogram (kg), and in the two weeks of eating a diet containing no ultra-processed food, they lost 1 kg. The trial confirmed the association indicated by observational studies. It also indicated a mechanism: in this case, excessive calorie consumption. Similar clinical trials have been carried out in other countries.
Your group follows thousands of Brazilians through the NutriNet Brasil study to observe how diet influences health. What have you found out?
We have been monitoring 30,000 people for four years. That includes young and old people, white and Black, rich and poor. We created one score to see how much ultra-processed foods people eat and on how many days of the week, and another score to track how much rice, beans, fruits, vegetables, meat, etc. they consume. In one of the studies, we showed that people who consume more than five ultra-processed foods per day are more likely to be overweight. Another study linked ultra-processed food with a higher risk of hypertension. These two studies suggest that the protective effect of healthy eating is partially lost when people continue to consume ultra-processed foods. A third study indicated a 40% increase in the risk of developing hypertension.
How has the consumption of ultra-processed foods changed in the last decade?
It has been increasing worldwide. When you analyze sales, it’s on the rise everywhere. In China, it’s scary. Ultra-processed food represented 1.5% of daily calories 15 years ago, and now it is 5%. In Latin America, it is also rising. In Brazil in particular, we are seeing a slower increase than before. Here, 20% of the population consumes ultra-processed foods.
What can we do to reduce consumption?
At the beginning of March, the federal government approved the new list of products that will be in the basic food hamper. It includes only minimally processed foods and culinary ingredients. These foods are also part of a list that the Brazilian Ministry of Health is drawing up to be exempt from taxes. The subject is yet to be discussed by the Office of the Chief of Staff before being sent to Congress for analysis.
Are ultra-processed foods cheaper than other foods in Brazil today?
It depends on the category. They are becoming cheaper. In general, ultra-processed foods are still more expensive. In the USA, for example, where they are already being produced en masse, changing things would require huge subsidies for natural and minimally processed foods. Here, tax reforms could help make things cheaper. Other options include restricting advertising for ultra-processed foods, supporting the traditional food system, and family farming, which provides food for public schools.
Is it ever okay to consume ultra-processed foods?
Yes, in two situations. One is the use of baby formula to feed infants aged up to 6 months when the mother cannot breastfeed, and the other is parenteral feeding for people who have had surgery. This is one of the differences in relation to tobacco—there is no situation in which smoking is recommended.
