A recurring theme on this site has been on the health effects of ultra-processed foods. The Twinkie and a Coca Cola represent the ultimate in ultra-processed foods. The NOVA classification of foods defined ultra-processed foods. Do these products represent the greatest dietary danger to American health? Or are the dangers overexaggerated?
Note, this post was inspired by an article by Linn Steward on Gourmetmetrics. In turn, a discussion between Linn and me posted on this site inspired her post. One of our discussion points was how a process differs from an ingredient. To a food scientist the two are distinct. To the home cook they are indistinguishable in the cooking experience. I illustrate my perspective by comparing a homemade twinkie with the real thing.
How do they make Twinkies in a processing plant? A food process is a series of steps to convert raw materials into finished product. A real Twinkie goes through seven steps as shown in this YouTube video:
-
- Mix ingredients into a cake batter and pour into pan molds,
- Bake the cakes,
- Inject the creamy filling into the baked cakes,
- Allow the filled cakes to cool as they become light and airy,
- Flip the cakes out of the pan once they cool,
- Separate the individual cakes into rows and wrap in plastic, and
- Pack into boxes for shipment to warehouses and then stores.
These snacks can be bought at many nearby stores or ordered online. They don’t last forever, but they have an estimated shelf life of 6 months. Each cake provides 16g of added sugar. The serving size is two Twinkies.
How can we make a twinkie at home? The instructions to bake a food at home are analogous to the steps in a commercial food process. The New York Times published a home recipe, and Linn Steward forwarded it to me. The steps it takes to make a twinkie at home include:
1. Preheat the oven to 350ºF,
2-3. Form molds from aluminum foil into 12 troughs,
4-7. Mix all ingredients to form the cake batter,
8. Scrape batter into molds and bake the cakes,
9. Mix ingredients together to prepare the cream center,
10. Remove the cakes from the molds, poke holes in the bottom, and insert cream.
The steps seem similar in the two processes until wrapping the commercial Twinkies in plastic and putting them into boxes. The shelf life of these tasty twinkies is less than a day as the cake absorbs the cream. The homemade cake contains about the same amount of sugar as a real Twinkie. Should that be considered added sugar if it is homemade? I do not consider the packaged Twinkie to be any more processed than the homemade twinkie. NOVA classifies the packaged Twinkie as ultra-processed. Should we classify the homemade twinkie as ultra-processed too?
But ultra-processed food was never really about processing. Ultra-processed has a nice ring to it—like a marketing term to sell an idea. The purported distinction in NOVA is the extent of processing a food undergoes. Thus, an ultra-processed food (Group 4) is more processed than a processed food (Group 3). A processed food undergoes more processing than a culinary ingredient (Group 2). Group 2 foods are more processed than Group 1 foods. Unprocessed and minimally processed foods make up Group 1.
By any understanding of food processing, NOVA is wrong. Fractioning, boiling, drying, and pasteurization are all considered minimal processes (Group 1). The degree of sophistication of these processes makes no difference. For example, ultra-pasteurized milk is not an ultra-processed food! Refining and milling are permitted processes for culinary ingredients (Group 2). Think table sugar and white flour. A sugar added at home or in a restaurant is fine as a culinary ingredient. The same sugar added in a processing plant drops a product to Groups 3 or 4. Canned, salted, sweetened foods find themselves designated as processed (Group 3).
Group 4 is about foods with additives. Forbidden processes include extrusion, molding, pre-processing for frying, and distillation. NOVA tells us that these processes have no “domestic equivalents.” Such comments defy the basic principles of food processing as described in Food Processing Technology. Pasta is an extruded product whether made at home or in a processing plant. Many homemade cookies, candies, and the homemade twinkies get their shape in molds. Breading of chicken before lowering in a kitchen deep-fat fryer is a domestic practice. Distilled spirits are Group 4, but not wine and beer. Home distilling is discouraged by many law enforcement agencies. The idea that pasta undergoes more processing than table sugar makes no sense. Likewise, chocolate kisses do not experience more severe processing than canned asparagus. Ultra-processing is all about using “industrial” ingredients.
What ingredients are present in the industrial-strength Twinkie? Twinkie ingredients appear on the package label. In his classic book Twinkie Deconstructed Steve Ettlinger describes the many ingredients in these snack cakes. Then he traces each ingredient back to its source. The makers of Twinkies have not reduced the number. The label of ingredients on the box that arrived from Amazon yesterday reads
Some obvious industrial ingredients include vitamins like thiamin mononitrate. Also present are dextrose (glucose), sodium acid pyrophosphate, and potassium sorbate. Note also xanthan gum, polysorbate 60, and two artificial colors. Each ingredient performs at least one important function. These industrial ingredients are all available at Amazon.com. Are they really industrial ingredients if they are available online?
How do those ingredients compare with those in the homemade twinkie? There are similarities, and there are differences. First, only the cook knows what is in the homemade twinkie as there is no label. If there was a label, the Ingredient Statement would read something like this
EGGS, SUGAR, MARSHMALLOW FLUFF (CORN SYRUP, SUGAR, WATER, EGG WHITES, ARTIFICIAL FLAVOR, XANTHAN GUM, LACTIC ACID, SALT), CAKE FLOUR, ALL-PURPOSE FLOUR, UNSALTED BUTTER, HEAVY CREAM, MILK, BAKING POWDER, VANILLA EXTRACT, CREAM OF TARTAR, SALT CONTAINS EGG, MILK, WHEAT
The homemade twinkie label doesn’t look that different from the real Twinkie. The twinkie has 15 ingredients; the Twinkie has at least 40. Only artificial flavor and xanthan gum look like industrial ingredients. Add the ingredients of the flours and baking powder in parentheses. The list becomes longer and more industrial ingredients appear. If we drop the Marshmallow Fluff from our recipe, the ingredients look more familiar. Without Marshmallow Fluff we have no cream filling. Without cream filling we have no twinkie.
Ultra-processed foods are really about ingredients. NOVA separates processed from ultra-processed foods by ingredients not by processes. At first, any processed food with more than five ingredients was an ultra-processed food. The concept is one we have heard before. Michael Pollan’s Rule #6 in Food Rules states “Avoid foods that contain more than five ingredients.” After getting pushback on the five-ingredient rule, NOVA proposed a new definition. Note that only distilled spirits lost their designation as ultra-processed foods. Almost all food additives, qualify a food as ultra-processed. The FDA considers all sugars in Group 2 as added sugars, but NOVA does not. NOVA forms the current basis of nutritional labeling in Brazil.
Which is better—the homemade twinkie or the ultra-processed Twinkie? That is a difficult question. I don’t see that much difference in the processes used to make them. Both versions involve mixing, molding, baking, filling, and wrapping. Their ingredients are similar, although the ultra-processed version has more industrial ingredients. They are both high in sugar and very sweet. I suspect that the fresh-prepared twinkie tastes better than the packaged one. The quality advantage appears to be short lived. When the cake starts absorbing the cream filling the next day, it begins to lose its appeal.
I can’t imagine that either one is healthier than the other one. Both versions have sugar, but the Twinkie contains high fructose corn syrup. Does that make the homemade twinkie healthier? The short shelf life of the homemade version could lead to overeating. Home baking likely leads to more labor and expense than buying the packaged version at a nearby store. For people who love to cook, the home experience sounds intriguing. It might make a great experience for teaching kids how to cook! For someone looking for a quick snack, the Twinkie would be the preferred choice. An earlier post found a similar comparison for homemade and store-bought mayonnaise.
Back when I ate foods containing gluten in them, I never cared for Twinkies. I doubt that I ever consumed five in my entire life. Now, peanut butter cups present an entirely different hazard for me. Ultra-processed foods do not undergo ultra-processing. Ultra-processed foods contain specific ingredients that keep them safe, extend their shelf life, and help prevent food waste. The five-ingredient rule persists to this day. There is nothing wrong with home cooking. There is nothing wrong with making foods more convenient for modern lifestyles. There is nothing wrong with an occasional sweet in our lives. Enjoy your favorite sweet treat, homemade or processed. But don’t overdo it.
Next week: The Thirteen Preservations by Allan Griff
In Defense of Processed Food 
Three cheers for discussions between folks who may not see things the same way but communicate based on mutual respect and the willingness to learn. I picked the following five comments on various aspects of processing / ultra-processing for response.
• Do these products represent the greatest dietary danger to American health?
No, probably not. But that doesn’t mean that a pattern high in foods classified as ultra-processed doesn’t contribute to an unhealthy pattern. Especially when these products replace freshly prepared meals, eaten at table, and shared with friends or family. I’ve finally developed an algorithm to assess my own collection of recipes based on the NOVA groups. I can now track calories by weight and by percentage for each of the four groups. I use ultra-processed foods everyday when I cook. Some days more than others of course but never even close to the over 50% as it’s reported that some of my fellow Americans do.
• Or are the dangers exaggerated?
Unfortunately I believe dangers have been exaggerated. And I’m upset because I believe exaggerations muddle the issue. Who is to blame? Here’s how I would allocate blame. First activists, no doubt with the best of intentions, have cooped a classification tool as a weapon to attack the industrial food industry. Most of these reformers, food writers, consultants, and health providers don’t even bother to make a distinction between traditionally processed and ultra-processed foods. Let’s call this group the destructive do-gooders. Second social influencers and marketeers who prey fears and sell product or services. Let’s call this group the charlatans. Consumers would benefit from a healthy dose of applied critical thinking because both destructive do-gooders and charlatans deserve scrutiny.
• Should the sugar added to the sponge mixture be considered “added sugar” if Twinkies are made at home?
Absolutely yes. Added sugar is a complex concept. It helps to step back and think about sugars from a processing perspective. Some sugars are intrinsic to a food and some sugars are added to food. The perspective works for intrinsic versus added fibers too.
• I do not consider the packaged Twinkie to be any more processed than the homemade Twinkie.
I do consider the packaged Twinkies to be ultra-processed. But I understand why from a food scientist’s perspective the lines of demarcation are squishy. I could not turn out 1000 Twinkies per minute in my home kitchen. That speed requires sophisticated industrial processing systems. However, as noted in the commentary, pasteurized and ultra-pasteurized milk are classified as minimally processed and I’m sure that assembly line runs a lot faster than anything I could do in my kitchen. So I don’t have a good answer. At least not yet.
• I suspect that the fresh-prepared Twinkie tastes better than the packaged one.
Me too. My best and most persuasive argument for NOVA is based on taste. I believe as research continues to explore the impact of industrial processing on our food supply, the role of advertising, and the social disconnect that snacking enables, that mechanisms will be identified and evidence will be forthcoming on the physical and mental health consequences of a dietary pattern based on an excessive amount of ultra-processed foods. But for now, I know that my opinion is a belief based on a leap of faith. So I’m left with taste. There are real and verifiable taste differences between a homemade food and an industrial analog.
LikeLike
Thanks again for your comments. Just two quick responses to your last two points.
I fail to see what speed has to do with anything. When the steps taken in a home and a processing plant are similar, the processes are similar. To somehow claim that a product of a mass process in a plant is somewhat more of a process than one in a kitchen makes no sense to me who has worked both in a plant and a home kitchen.
Yes, the home cake done correctly probably tastes better than the commercial product right out of the oven. But when it becomes a mass of mush the day after it is baked as the cream filling soaks into the cake, it will no longer taste better. Are all of the 12 homemade twinkies consumed on the day of baking leading to a massive intake of added sugar that day? Or are the leftover cakes, now a soggy mess thrown away?
LikeLike
twinkie twinkie, little star, do you wonder why you are
processed or ultraprocessed? keep firing – especially with a Griff message
LikeLike
dear noah webster shewfelt
now you is an etymologist?
LikeLike