Do ultra-processed foods increase Parkinson’s risk?

More than half of the average American adult’s calories come from ultra-processed foods, which have minimal nutritional value and often contain artificial additives. A growing body of research suggests that eating large amounts of such foods may raise the risk of various chronic diseases. New evidence links high consumption with greater likelihood of developing features that precede Parkinson’s disease.

Researchers at Fudan University in Shanghai, China, in collaboration with Alberto Ascherio at the Harvard T.H. Chan School of Public Health, worked with the self-reported diet data from more than 40,000 health professionals, tracked since the mid-1980s. In the 2010s, the participants also answered questions about early nonmotor features associated with Parkinson’s.

Comparing these data points, the team found that people who consumed about 11 servings of ultra-processed foods per day were 2.5 times likelier to develop early nonmotor features than people who consumed about two to three servings.

In this edited conversation, Ascherio explains the significance of the finding and what’s still unknown.

---

Going into this study, what’s the understanding about how diet can affect brain health, particularly neurodegenerative diseases like Parkinson’s?

There’s no question that diet is important. In fact, there’s a consensus that we could delay the onset of many neurodegenerative diseases by being physically active and having a healthy diet. But there is still a lot that we need to understand, particularly when it comes to Parkinson’s, which is a bit of an exception among chronic diseases. For instance, cigarette smoking is associated with a lower risk of Parkinson’s — it’s one of the very, very few diseases in which cigarette smoking doesn’t raise the risk of the disease. With diet, it’d be nice if the diet that delays cardiovascular disease or cancer also contributes to prevent Parkinson’s. But we don’t know yet.

We do know a few things. We’ve previously found that caffeine intake is associated with a lower risk of Parkinson’s disease. There’s evidence that the Mediterranean diet, which is associated with lower risk of many chronic diseases, might do the same for Parkinson’s, but it’s not as clear as with cardiovascular disease or diabetes. And we’ve seen that intake of flavonoid, which is a component of fruits and vegetables, also tends to be modestly associated with lower risk of Parkinson’s. I think we’re still missing something. In this study with ultra-processed food, though, we found much stronger signals than we’ve seen before.

Why did you focus on early symptoms, as opposed to people who developed Parkinson’s? And why ultra-processed food?

Parkinson’s disease really starts at least 10 to 15 years before the diagnosis. If you study individuals with diagnosed Parkinson’s disease, you’re really 15 years late. So we wanted to find out which factors may influence the development of Parkinson’s disease. For this purpose, we created this large cohort in which we assess the presence of what we call prodromal features of Parkinson’s disease, which occur many years before the typical signs and symptoms of the disease.

In terms of the dietary aspect, we’ve been studying nutrient and dietary patterns for many years, but more recently, it became of interest to look at food in terms of the amount of processing and artificial ingredients. Food processing has always been a necessity for preservation and shelf life, but we’re talking about something different with ultra-processed food. This is really an industrial process to make the food more attractive. It goes well beyond what is needed to preserve the food and is more related to the marketing and commercial aspects of the food industry.

Which prodromal features, or early signs, did you focus on?

We had a huge cohort, so we selected from the clinical literature those common prodromal features that are relatively easy to assess with just simple questions or inexpensive tests. One feature was acting out dreams, which we only asked of people who had a sleeping partner who could verify it. Another was hyposmia, or reduced sense of smell. For that, there was a scratch-and-sniff test that we mailed out for people to smell. And also intestinal constipation, a non-specific feature that can be significant when combined with other signs.

Obviously when you look at these features one by one, they are very common, but the combination of three or more is present in only about 2 percent of older adults. We’ve found in past research that people with all three of these features are 23 times more likely to develop Parkinson’s disease.

“Will eating less of these foods be sufficient to prevent Parkinson’s? Almost certainly not, but it could be beneficial. So far, the strongest evidence for reducing Parkinson’s risk is for physical activity.”

You found an association, but not necessarily a cause. How likely is it that the ultra-processed food is causing the development of the prodromal features?

It is very challenging. There’s always the risk of reverse causality, where the prodromal features influence the diet, rather than diet influencing the prodromal features. For instance, if you’re constipated, you might eat more fruits and vegetables to offset the constipation. But we tried to address this by looking at diet very early on. We started assessing the prodromal features in 2012, and we found that there’s already an association between those features in 2012 and the diet back in 1986. In other words, if you have high consumption of ultra-processed food in 1986, you’re more likely to have these prodromal features in 2012. If it was reverse causation, you wouldn’t expect an association to be present that early.

The other limitation is what we call confounding: Could the consumption of ultra-processed food be a marker or something else? We try to adjust for other factors, but the adjustment is never perfect or complete. We do not know yet whether there’s something toxic in the ultra-processed food, or maybe, given that they contribute to 50 percent or more of your calories, you’re just missing other components of the diet that are protective.

If ultra-processed food is the cause, what might the mechanism be?

We don’t really know. But there is evidence that some chemicals can cause Parkinson’s disease. The evidence is particularly strong for pesticides and herbicides. It’s clear that they can increase risk; we just don’t know which pesticides do that. This is because we are exposed to combinations of pesticides and herbicides, not to individual compounds.

Here, we are in a similar situation. We can see that this ultra-processed food is associated with a higher frequency of these prodromal features, but we really don’t know which component. And, as I mentioned, it could even be a displacement of nutrients rather than a specific toxic effect.

What should people do with this information?

Well, given the accumulated evidence related to risk of cardiovascular disease, diabetes, chronic decline, and dementia, there’s no question that we should limit consumption of ultra-processed foods as much as possible. And it seems that Parkinson’s is in the same direction. Will eating less of these foods be sufficient to prevent Parkinson’s? Almost certainly not, but it could be beneficial. So far, the strongest evidence for reducing Parkinson’s risk is for physical activity. I know it’s a bit disappointing to not have a clear answer. It’s a very difficult question to address, but it’s one that we’ll continue to study.

---

This research was funded in part by the NIH National Institute of Neurological Disorders and Stroke and the Department of Defense.