A couple of threads fell into place recently as I continue to explore the NOVA food classification system. And that send me out looking for Twinkies.
The first thread came by accident. I found myself listening to a previously recorded presentation at Google headquarters in California by New York based writer, editor, and publisher Steve Ettlinger. He was in California discussing his book, Twinkie, Deconstructed: My Journey to Discover How the Ingredients Found in Processed Foods Are Grown, Mined (Yes, Mined), and Manipulated into What America Eats. Published in 2007, the book is a fascinating exploration of the multitude of weird, unrecognizable ingredients used in modern food processing.
The second thread fell into place when I discovered a recipe for homemade Twinkies does actually exists. About a decade ago, Jennifer Schihauser, a NYT journalist, developed a series of articles around the concept “If I make it myself, it’s not junk” and one of her recipes was for making Twinkies at home.
With a Twinkies formulation in one hand and a recipe for a homemade version in the other hand, I can actually do an “apples to apples” comparison between a formulated product and an equivalent homemade version. That was the third thread falling into place. I have what I need to count ingredients / markers of Ultra-Processed food (UPF) and compare nutrition stats.
I needed to go out looking for Twinkies to get a label, always the best way to get the ingredient list. Product formulations change so the best source of data is always the current product label. Besides, I’ve never tasted a Twinkie so I was curious. I figured I should at least have a bite.
Using the ingredient list from the Twinkie’s wrapper and the Jennifer Schihauser recipe, I was ready to begin.
Running the numbers was easy because I’m in the business of recipe analysis. For guidance on the NOVA food classification system, however, I selected 2019 commentary Ultra-processed foods: what they are and how to identify them to serve as my reference guide.
INGREDIENTS & UPF MARKERS
Flour. Butter. Sugar. Eggs. Those are the primary ingredients listed in the homemade version. All familiar, recognizable ingredients. Besides the basics, the recipe also calls for salt, milk, vanilla, cream of tartar, heavy cream, and Marshmallow Fluff. That’s a total of 11 ingredients.
The formulated Twinkie is manufactured with mostly unfamiliar ingredients. To my eye, they look like deconstructed fragments. Examples are high fructose corn syrup or dextrose or hydrogenated tallow, mono and diglycerides, polysorbate 60, soy lecithin.
I counted 7 primary ingredients plus an additional 22 listed on the label as <2% by weight and felt comfortable tagging 18 as UPF markers.
What counts as a marker? As per my NOVA reference document, markers can be 1) “food substances not usually found in the home kitchen like high-fructose corn syrup or 2) “cosmetic additives” like flavors, colors, emulsifiers, sweeteners, or thickeners. Note that salt and sugar get a pass because NOVA classifies them as processed culinary ingredients.
Both homemade and formulated versions are intense concentrations of fats and sweetness with little positive nutrient benefit. And both versions are calorie dense.
Nutrients per serving size is the best metric for checking how much you’re actually eating, but comparing nutrients in similar products needs a different metric. My choice for this type of comparison is the calorie density scale which reflects nutrient values by weight. Think of the calorie density scale as a sort unit price for nutrients.
The homemade Twinkie is 3.6 calories per gram. The formulated product is 3.4 calories per gram. To put these calorie density numbers in context, Doritos scores 5.1 on the calorie density scale whereas a Chipotle Beef Burrito will clock in at about 1.8 calories per grams.
Both recipe and formulation have comparable amounts of total fat but homemade Twinkies have a significantly higher concentration of saturated fat. No surprises here. Butter is the only fat called for in the recipe and butter is notoriously high in saturated fatty acids.
The formulated Twinkies use a combination of animal fat (tallow) and seed oil (cottonseed oil) which results in less saturated fat and more unsaturated fat.
Both recipe and formulation are intensely sweet. Sweeteners in one form or another comprise about 40% by weight. Homemade uses granulated sugar plus some marshmallow fluff. The formulation lists sugar, high fructose corn syrup, and dextrose.
Sodium is not significant for either version but it’s worth noting there’s more in the formulation than the recipe.
TASTE IS 100% SUBJECTIVE
The final thread fell into place when I tasted the Twinkie. Actually better than I expected but still too sweet for my taste.
It’s takes many hours to source and study the NOVA research, to count ingredients and identify the markers, and to compare nutrition stats. Here’s why I believe my time was well spent:
- I learned that nutrients are easier to count than ingredients or UPF markers, especially hard was deciding which additives were “cosmetic” and which were not. Difficult and inconclusive.
- Simplistic pronouncements by food pundits, corporate marketeers, or social influencers are misleading. NOVA is a sophisticated concept. Selecting a couple of buzz words may sell books or products but does a disservice to consumers.
- Expanding my NOVA analysis skills is a good investment of time. My sense is we’ll be having more serious discussions about degree of processing in the coming years.
So do I plan to do more apples-to-apples comparisons between a formulation and a recipe? You bet I do.