Online poll shows Americans can’t tell healthy processed foods from unhealthy ones

Thirty-nine percent of respondents to an online poll incorrectly said all processed foods are unhealthy, according to new research by the Physicians Committee for Responsible Medicine published today, July 8, 2025, in JAMA Network Open. When asked which foods increase type 2 diabetes risk, 51% cited sugar, 19% said desserts, and 7% said carbohydrates, in general—all inaccurately.

The Physicians Committee/Morning Consult survey was conducted online among 2,174 U.S. adults Dec. 13 to 15, 2024.

“Foods considered ‘ultra-processed’ can reduce diabetes risk or increase it, depending on which ones they are,” said Neal Barnard, MD, lead author of the paper and president of the Physicians Committee.

“Studies show it’s processed meat consumption that is associated with diabetes and cardiovascular disease, not plant-based ultra-processed foods, such as breakfast cereals, which are actually associated with reduced risk of these conditions.”

The news comes on the heels of a new American Medical Association policy supporting public awareness and education about the differences between healthful and unhealthful ultra-processed foods, and just before the federal Food and Drug Administration (FDA) is expected to release a Request for Information for stakeholder input in defining ultra-processed foods.

Though the four-tiered NOVA system developed by Brazilian researchers in 2010 has commonly been used to define ultra-processed foods, the FDA has not formally developed its own definition, which could eventually be used in the next Dietary Guidelines for Americans. The guidelines are used to set the nutritional standards of the National School Lunch Program and other federal food assistance programs.

“It is important that a federal definition of ultra-processed foods reflect scientific evidence and avoids aggravating public confusion,” Dr. Barnard said. “Science shows that ultra-processed foods differ greatly: Certain ones are associated with health problems while others are associated with reduced risk.”

A Harvard University study showed that frequent processed meat consumption was associated with a 44% increased risk of diabetes, while consumption of ultra-processed cereals had the opposite effect—a 22% reduction in diabetes risk.

A recent BMC study showed moderate consumption of muesli and bran cereal correlated with reduced mortality risks related to cardiovascular disease and cancer.

Animal products are a major source of saturated fat in the American diet. A diet high in saturated fat—not carbohydrates—has been shown to reduce insulin sensitivity in as little as four weeks in individuals with a normal insulin sensitivity at baseline, despite no changes in their body weight.

As to the attitudes of U.S. adults, a 2023 survey showed similar results to the recent survey. Respondents’ perceptions of the unhealthfulness of foods correlated strongly with “processing” levels. In the older survey, most young participants failed to identify processed meat as a “processed food.”

And while consumption of both processed and unprocessed meat is associated with diabetes (and colorectal cancer and cardiovascular risk), most young participants did not identify these products as increasing the risk of developing type 2 diabetes, even given the rising incidence of diabetes in this group.

“The vague term ‘processed foods,'” Dr. Barnard said, “should be replaced by more specific terms describing foods’ known health effects. Some plant-based foods that are considered processed, such as cereal and canned vegetables and fruit, are actually what Americans need to fight chronic lifestyle diseases like diabetes and cardiovascular disease.”

Provided by
Physicians Committee for Responsible Medicine