The American Heart Association published a science advisory today in Circulation on ultra-processed foods — industrial formulations built from refined ingredients and additives, the category the Nova classification calls group four. Its conclusions: the association between high intake and heart attacks, strokes, type 2 diabetes, obesity and early death is consistent and dose-dependent; most such foods are high in saturated fat, added sugar and salt; a small number, like some whole-grain breads and low-fat dairy, can fit a healthy pattern; and keeping the high-fat, high-sugar, high-salt kind below roughly a tenth of daily calories is a prudent target while the science is refined. More than half of the calories Americans eat now come from this category, and more for children.
What the advisory does not claim. The honest part of this document is its list of gaps. The association is strong and observational; whether processing itself harms you beyond the sugar, fat and salt it carries is not established, and the authors say so. They call for trials comparing nutritionally matched processed and unprocessed foods and for the FDA to modernize how it evaluates additives. The one controlled trial that exists points the way: in a 2019 NIH study, people fed ultra-processed meals matched for calories and nutrients ate about 500 more calories a day than on whole foods, without noticing. Something about the food, not just its nutrients, overrides the signal that says stop.
My patients do not need the Nova categories. They need one question: can you picture every ingredient on the label? A loaf of bread with five ingredients passes. A snack bar with thirty, half of them unpronounceable, does not. The advisory's tenth-of-calories target is useful for a researcher. For a person standing in a grocery aisle, "would your great-grandmother recognize this as food" is the same rule, and it works without a chart.
The industry's answer is already written: "healthy ultra-processed foods," reformulated with a halo. The advisory gives it an opening by naming a small category of exceptions, and the marketing will drive a truck through it. Watch for the studies that follow — who funds the ones arguing that processing is harmless, and who funds the ones arguing it isn't. The funding line is where this debate will actually be fought.
Why it matters to a recovering optimizer. The nutrition chapter of the book we are writing makes one argument: the specific diet matters far less than whether it is made of real food. Today the country's largest heart organization said, in the measured language of a science advisory, roughly the same thing. You do not need macros, timing or a tribe. You need the half of your calories that comes from a factory to come from a kitchen instead.
Don't memorize the Nova categories. If you can't picture every ingredient, it's the category.