Nutrient density and processing answer different questions

Define the two ideas before comparing

Nutrient density has no single universal denominator unless you specify it. A comparison per 100 grams, per serving or per calorie can produce different rankings. It also depends on which nutrients you include and whether the food database has reliable values for each one. FoodData Central is a useful reference for named foods, but a package’s actual formulation, portion and cooking still matter.

Processing is another dimension. NOVA groups foods by the nature and purpose of processing, including industrial formulations. That can help people notice patterns that a basic nutrient panel misses. It does not calculate vitamin and mineral breadth, and it does not mean every item in the same NOVA group has the same nutritional profile.

A useful comparison keeps food role visible

Compare an apple with a serving of eggs: the apple can be a useful fruit snack, while eggs contribute protein and a different range of nutrients. Calling the apple “less nutrient-dense” in some dimensions does not make it a bad food; it tells you that the two foods play different roles. A meal can use fruit for one role and eggs, fish, meat or tolerated dairy for others.

Now compare plain Greek yoghurt with a highly formulated protein dessert. Both may provide protein. The ingredient list, added sugars, processing markers, portion and wider nutrient panel answer additional questions. A high protein number alone does not erase the rest of the formulation; a longer label alone does not prove harm.

Why one score cannot be the whole diet

A product score compresses multiple judgments into one number. It can help you compare similar products quickly, but it cannot represent your portion, allergy, budget, preferences, meal pattern or lab results. It is especially misleading to call a score a “micronutrient calculation” when a record only contains calories and macronutrients. Missing fields remain unknown, not zero.

Optimally’s score is a curated editorial rule set informed by food composition and its stated Primal-inspired approach. It is not clinically validated, does not diagnose deficiencies and does not establish that a product prevents or treats disease. Whole-food status does not automatically earn the top rating; category calibration considers a food’s role and the available nutrition information. The current apple maximum is 90, which reflects that calibration rather than a harmful-food warning.

A four-question comparison routine

• What job is this food doing: meal protein, fruit, cooking ingredient, snack or convenience item?
• What does the label actually report, and for which serving? Separate declared values from reference database values and qualitative descriptions.
• What does the ingredient list say about formulation and processing? Look for the specific ingredient you care about instead of relying on a “clean” front claim.
• What else is on the plate across the day? Variety and enough energy matter more than finding one perfect product.

Processing evidence has a scope

In a small 2019 inpatient randomized crossover trial, NIH researchers fed adults ultra-processed and minimally processed diets for two weeks each. Participants ate more calories and gained weight during the ultra-processed phase even though the offered diets were matched on several nutrients. That is meaningful experimental evidence about those diets and those participants; it does not prove that every processed product is harmful, explain every mechanism, or validate every scoring system.

Use processing as one useful lens alongside nutrient content and personal context. That gives you a practical choice without turning food into a purity contest.

Common questions