In “The Problem with Processed Food” I point out 3 problems with highly processed food:
Most food processing makes food easier to eat and enhances digestibility [which can increase the speed with which it boosts insulin].
The newer the type of food processing, the less tested it is by time.
Food companies have a different objective function than you do—or at least a different objective function than your long-run self does.
Under the category of things “untested by time” is propionate. The article shown above reports on recent research:
Consumption of propionate, an ingredient that’s widely used in baked goods, animal feeds, and artificial flavorings, appears to increase levels of several hormones that are associated with risk of obesity and diabetes …
The study, which combined data from a randomized placebo-controlled trial in humans and mouse studies, indicated that propionate can trigger a cascade of metabolic events that leads to insulin resistance and hyperinsulinemia — a condition marked by excessive levels of insulin. The findings also showed that in mice, chronic exposure to propionate resulted in weight gain and insulin resistance.
Additional details strengthen the story—except for the limitation of the human trials to short-run effects on only 14 people. However, that was enough to show that the mouse results seem relevant to humans:
For this study, the researchers focused on propionate, a naturally occurring short-chain fatty acid that helps prevent mold from forming on foods. They first administered it to mice and found that it rapidly activated the sympathetic nervous system, which led to a surge in hormones, including glucagon, norepinephrine, and a newly discovered gluconeogenic hormone called fatty acid-binding protein 4 (FABP4). This in turn led the mice to produce more glucose from their liver cells, leading to hyperglycemia — a defining trait of diabetes. Moreover, the researchers found that chronic treatment of mice with a dose of propionate equivalent to the amount typically consumed by humans led to significant weight gain in the mice, as well as insulin resistance.
To determine how the findings in mice may translate to humans, the researchers established a double-blinded, placebo-controlled study that included 14 healthy participants. The participants were randomized into two groups: One group received a meal that contained one gram of propionate as an additive and the other was given a meal that contained a placebo. Blood samples were collected before the meal, within 15 minutes of eating, and every 30 minutes thereafter for four hours.
The researchers found that people who consumed the meal containing propionate had significant increases in norepinephrine as well as increases in glucagon and FABP4 soon after eating.
Although I have recommended that the first thing you should do to make your diet healthier is to go off sugar (see for example “3 Achievable Resolutions for Weight Loss” and “Letting Go of Sugar”) I have been careful to point out that as things stand it is hard to distinguish between going off sugar and going off highly processed food, because almost all highly processed food has sugar as an important ingredient. There really could be reasons other than sugar that make processed food bad for you, but if you go off sugar, you will avoid those problems too, because you will be going off most processed food as well.
One claim I have made is that an insulin spike will make someone hungry again. Note that for the producers of processed food, that would typically seem like a plus: there is a reasonable chance that the person eating their food would decide that they were hungry for that particular food, or that, being hungry, it was again most convenient to eat that food. So even if the producer of that processed food were only looking at the sales numbers, they might gravitate toward producing types of processed food that caused insulin spikes. Thus, even without any conscious bad intent, the process of engineering food that people will eat a lot of has some tendency to result in bad outcomes. And that engineering of food happens much, much faster than any human evolution that could blunt bad effects.
Michael Pollan famously said “Eat food, not too much, mostly plants.” This post on Michael Pollan’s statement gives several criteria for something being real food instead of just a food-like substance. If it is real food:
Your great grandmother would recognize it as food.
It doesn’t come out of a container with a long list of ingredients.
It will rot or go bad, and so tends to be on the outer perimeter of grocery stores where it is easier to bring in fresh food of that type.
The biggest two limitations of Michael Pollan’s dictum is that (a) not all plant food is created equal in terms of its health effects, and (b) eating all the time, from morning til night, is problematic. See “Forget Calorie Counting; It's the Insulin Index, Stupid,” “Stop Counting Calories; It's the Clock that Counts” and “Jason Fung's Single Best Weight Loss Tip: Don't Eat All the Time.”
For annotated links to other posts on diet and health, see: