Ultra-processed foods are part of daily life, filling 60% to 70% of the American diet. That umbrella term covers a wide variety of packaged foods, from snacks carrying added sugar, salt, and saturated fat combined with preservatives for shelf life, additives for flavor, colorants for eye appeal, and chemicals like bisphenols in the packages themselves. Processed meats, as their name would suggest, also belong in this category.
But ultra-processed foods are not all the same. Whole grain cereal and breads also qualify for the label, as do yogurts and some other dairy desserts. JoAnn Manson, a professor of medicine at Harvard Medical School and Brigham and Women’s Hospital, reported in Lancet Regional Health last month on a large analysis of long-running studies looking at links between ultra-processed foods and cardiovascular disease. The bottom line: One size does not fit all.
Aiming to fill a gap in what’s known about these foods, the researchers drew on data from three prospective cohort studies of more than 200,000 nurses and doctors over three decades as well as a meta-analysis of 22 international studies with 1.2 million participants. They found a significant increase in coronary heart disease (23%), overall cardiovascular disease (17%), and stroke (9%) in people who said they consumed the most ultra-processed foods compared to those who said they ate the least.
Looking more closely, some foods were more tightly tied to higher risk of disease: sugar-sweetened drinks and processed meat were linked to higher risk, while breakfast cereals, yogurt, some dairy desserts, and whole grains were associated with comparatively lower risk.
Manson spoke with STAT last week about the study’s conclusions, following a meeting of the Dietary Guidelines for Americans Committee in which recommendations on ultra-processed foods were not made, owing to a lack of strong evidence. This conversation has been edited for length and clarity.
The committee charged with making recommendations for U.S. dietary guidelines didn’t weigh in on ultra-processed foods because the evidence is lacking. Why is this such a tough question to answer?
It has been challenging to do this research because most of the diet assessments in different research studies do not really rigorously classify people by ultra-processed food intake. Randomized trials have to be very large, very long duration to get answers about prevention of chronic diseases. And so it’s just not realistic.
In the recent Lancet Regional Health publication, we had 1.2 million people in the meta-analysis, which is obviously huge. And that’s what it took to look at the relationship between ultra-processed food intake and risk of cardiovascular disease, and then to look at subtypes of ultra-processed foods, where we did see diversity. It was not one size fits all.
Were you surprised by what you found?
We actually were surprised that there were several types of ultra-processed foods linked to lower risk of cardiovascular disease. We did not expect as much diversity across types of ultra-processed foods.
Can certain types of these foods be part of a healthy diet?
I do think we have to accept there is a complexity and diversity in this association between ultra-processed foods and health. It’s important to focus on a whole-food diet, primarily plant-based. However, it does mean that certain types of ultra-processed foods can be incorporated into a healthy diet. They don’t need to be eliminated entirely. They don’t need to become taboo. We know that doesn’t work with dietary guidelines.
Some of these other foods that fall into the category of UPFs — the breakfast cereals, the yogurt, whole grain bread — these are not linked in these large epidemiologic studies with increased risk of cardiovascular disease. There can be a little latitude incorporating certain types of ultra-processed foods into a healthy diet and not eliminating them entirely. And then we’ll see how the evidence accrues over the next several years.
Should the next round of dietary guidelines reflect this?
I suspect that the ultra-processed foods will be in the 2030 dietary guidelines. I think it’s probably going to take another several years to have a sufficient evidence base. But I think it’s going in that direction and that more and more of the large cohort studies are able to look at the relationship with cardiovascular disease, cancer, diabetes, all-cause mortality, cognitive outcomes, depression — many of these really important health outcomes. I think we’re going to have a much larger evidence base at that time.
What can speed the process along?
The research is going in the direction of trying to look at more biomarkers of ultra-processed intake: not only insulin resistance, glucose tolerance, and change in weight, but also looking at the effect on the microbiome, the metabolomic patterns, the changes in the proteome, and then DNA methylation markers for epigenetic aging. All of these can be looked at in terms of chronic disease outcomes, and I think that will further accelerate the research in this area.
I think there are ways to do randomized trials. They’re probably not going to be like some of our trials in 25,000 participants. But it can be trials of 100 people looking over a few months at changes in some of these biomarkers and changes in weight.
What about alcoholic beverages?
More risks are being identified for even moderate intake of alcohol and cardiovascular outcome. Atrial fibrillation, for example, which is becoming so common, is linked to alcohol intake. And hemorrhagic stroke. There’s also concern about breast cancer and other forms of cancer. I think there’s likely to be increasing research on alcohol and health. And again, just as with ultra-processed foods, it may take a few more years to really have it incorporated more widely into guidelines. I think it’s moving in that direction.
We certainly should not be recommending to people who don’t currently drink that they begin to consume alcohol in order to lower their risk of heart disease. That is not likely to improve health and could have very adverse effects on health.
Do you see progress?
I am hopeful. I think that the research on dietary patterns has been very helpful. Identifying certain patterns that allow for flexibility, including some of the cultural influences and cultural differences, can really make a difference in not forcing people into these very restrictive types of recommendations.
I definitely agree with recommendations to reduce and minimize ultra-processed foods. And if you’re going to have them, you can lean more toward the ones that have been associated with lower risk or neutral associations with cardiovascular disease or other chronic diseases as opposed to the increased risk, particularly reducing sugar-sweetened beverages and processed meats.
Does the rising use of obesity drugs factor into any of these guidelines?
I think it’s too early to know. I think we need a lot more research on the long-term effects of the GLP-1s. And I still think the composition of the diet is going to be important. I think the answer is not going to be just throwing medications at these enormous chronic health problems. Much more of a focus on prevention is what we need.
STAT’s coverage of chronic health issues is supported by a grant from Bloomberg Philanthropies. Our financial supporters are not involved in any decisions about our journalism.