Most nutrition advice you see is some variation of: cut sugar, watch calories, eat more vegetables. That advice isn’t wrong. It’s just shockingly incomplete.
Here’s what the science actually says: a study published in BMJ in 2024, following 100,000+ adults for 10 years, found that the degree of food processing predicted cardiovascular disease, metabolic disease, and all-cause mortality more strongly than calorie intake, sugar intake, or saturated fat intake combined.
A diet of “low-calorie ultra-processed food” was worse for outcomes than a diet of “high-calorie unprocessed food.”
The framework researchers used to measure this is called NOVA classification. Almost no consumer-facing nutrition app shows it. We think that’s a problem.
What NOVA classification actually is
NOVA was developed at the University of São Paulo in 2009 by epidemiologist Carlos Monteiro and his team. It sorts every food into one of four categories based on how much it has been processed, not what nutrients it contains.
| Category | Description | Examples |
|---|---|---|
| NOVA 1 — Unprocessed or minimally processed | Whole foods, possibly cleaned, dried, or frozen | Apples, oats, eggs, raw chicken, plain Greek yogurt, milk |
| NOVA 2 — Processed culinary ingredients | Substances extracted from NOVA 1 foods, used in cooking | Olive oil, butter, salt, sugar, honey |
| NOVA 3 — Processed foods | NOVA 1 + 2 combined, recognizable as food, traditional processing | Cheese, bread (artisanal), canned tomatoes, fermented vegetables, smoked fish |
| NOVA 4 — Ultra-processed foods | Industrially formulated with substances extracted or synthesized in labs; high in additives | Soft drinks, packaged snacks, breakfast cereals, instant noodles, most “low-fat” or “diet” products, hot dogs, frozen pizza, flavored yogurts |
The line between NOVA 3 and NOVA 4 is the line researchers care most about. NOVA 3 is food. NOVA 4 is food-like industrial product.
How to spot NOVA 4 without a degree in food science
The fastest test: read the ingredient list. If you see any of these, it’s almost certainly NOVA 4:
- Words ending in
-ose(high-fructose corn syrup, dextrose, maltodextrin) but not in the natural fructose/glucose found in fruit - Hydrolyzed proteins (whey isolate, soy protein concentrate)
- Modified starches (modified corn starch, modified food starch)
- Hydrogenated or interesterified fats
- Emulsifiers (soy lecithin, mono- and diglycerides, polysorbate 80, carrageenan)
- Artificial sweeteners (aspartame, sucralose, acesulfame-K)
- Flavor enhancers (MSG, disodium inosinate, autolyzed yeast extract)
- Bulking agents (microcrystalline cellulose)
- Coloring agents named with E-numbers or “Red 40” / “Yellow 5” style codes
If the ingredient list reads like a chemistry exam, it’s NOVA 4.
The second fastest test: was the food made by hand 100 years ago? Bread, cheese, yogurt, jam, pickles, sausage — yes, all NOVA 3. Cheez-Whiz, Lunchables, Pop-Tarts, sugar-free Jell-O — invented after 1950, all NOVA 4.
Why NOVA matters more than calorie counts
This is the part most nutrition apps miss.
A calorie-balanced ultra-processed diet does worse than a calorie-balanced minimally-processed diet. Multiple studies, controlled for everything we know how to control for, including a randomized inpatient trial by Kevin Hall at the NIH (2019) where 20 participants ate ultra-processed and unprocessed diets matched for calories, sugar, fat, fiber, and macros. The same participants. The same total calories.
Result: on the ultra-processed diet, they spontaneously ate 500 calories more per day and gained 0.9kg in two weeks. On the unprocessed diet, same nutrient profile, they lost weight without trying.
The hypotheses for why:
- Hyperpalatability — ultra-processed food is engineered to override satiety signals
- Faster eating speed — UPF dissolves easily, you eat before fullness kicks in
- Disrupted gut microbiome — emulsifiers like polysorbate 80 thin the gut lining
- Endocrine disruption — some additives may interfere with hunger hormones (leptin, ghrelin)
- Fewer micronutrients per calorie — UPF is calorie-dense but nutrient-sparse, so the body keeps signaling hunger
The takeaway: a Coca-Cola and a homemade fruit smoothie with the same sugar content do not affect your body the same way. A “low-fat ultra-processed snack” is worse than a “high-fat whole-food snack” of equal calories. This is contrarian to 30 years of nutrition advice. The data is clear.
How NOVA shows up on your kitchen shelf
Take a typical North American breakfast: Frosted Flakes (Kellogg’s) with skim milk and orange juice from concentrate. By calorie count, that’s “balanced.” By nutrition label, it’s “fortified with 11 essential vitamins.”
Run it through NOVA:
- Frosted Flakes: NOVA 4 (ultra-processed — milled corn flour + sugar + malt flavoring + BHT preservative + iron added back after stripping)
- Skim milk: NOVA 1 (minimally processed)
- OJ from concentrate: NOVA 4 (ultra-processed — heat-treated, water removed, water added back, “natural flavors” reconstituted)
Two of three items: NOVA 4. The “vitamin fortification” on the cereal box doesn’t undo the processing damage. The ingredient list does.
Compare to: rolled oats with whole milk and a fresh apple. NOVA 1 + NOVA 1 + NOVA 1. Same calories. Different metabolic outcome.
This is exactly what generic nutrition apps miss. They look at sugar content, fat content, vitamin content. They don’t look at whether the substrate is recognizable as food.
What we built into SYE
When you scan a product in SYE, every ingredient is run through a NOVA classifier. The result shows up in the score and explanation. If you’re one of the (rapidly growing) audience trying to cut ultra-processed food specifically, you can also enable a “Avoid Ultra-Processed” filter in your dietary profile, and any NOVA 4 product will be flagged AVOID regardless of how its calories or sugar look.
This is genuinely useful for:
- People who’ve read about UPF and want to act on it — most apps make this hard. NOVA isn’t a marketing label, it’s a research tool.
- People with gut conditions (IBS, IBD, SIBO) — emulsifiers in NOVA 4 are a known irritant for many.
- People recovering from disordered eating — focusing on ingredient quality (NOVA) is healthier than counting calories or macros for many.
- Parents — toddler snacks marketed as “healthy” are often NOVA 4. Most parents don’t realize this.
How to read your shelves with new eyes
If this is the first time you’re hearing about NOVA, an exercise: walk into your kitchen, pick 10 packaged products, and read the ingredient lists. Categorize each as NOVA 1, 2, 3, or 4.
Most people are surprised by what they find. The “healthy” granola bars, the “natural” yogurt, the “all-natural” cereals — many of these are NOVA 4. Marketing has gotten very good at making industrial products look like artisanal food.
If you find your kitchen is 70%+ NOVA 4, you’re not unusual. The average North American diet is 58% NOVA 4 by calories (NHANES data, 2018). The change isn’t to throw everything out. The change is to gradually swap NOVA 4 for NOVA 1-3, item by item, as you run out and re-shop.
You don’t have to be perfect. The research suggests even moving from 60% to 40% NOVA 4 has measurable health effects.
What to do next
- Look at three things in your kitchen right now. Read their ingredient lists. Are any NOVA 4?
- When you next grocery shop, replace one item with a NOVA 1-3 alternative.
- If you want a tool that does the categorization automatically, download SYE and scan as you shop.
NOVA classification isn’t a fad. It’s the framework academic nutrition researchers actually use to study disease risk. Generic nutrition apps haven’t caught up yet because their scoring was designed in the 1980s when “fat-free” was the goal. The science has moved. Your apps should too.
If you want to read more, “Generic food scores fail diabetics” covers the related issue of why personalization in food scoring matters even more than absolute scores.