Nutrition panels are standardised, regulated and useful, and there is a category of relevant information they were never designed to convey.
What the panel covers
Energy, fat, saturated fat, carbohydrate, sugars, protein and salt in most regulatory frameworks, with variations by jurisdiction.
Fibre is required in some and voluntary in others.
Micronutrients appear where a claim is made or where fortification has occurred.
Which is a specific list rather than a complete description.
Calorie figures are approximations
Energy content is generally calculated using standard factors per gram of each macronutrient rather than measured directly.
Which is reasonably accurate in aggregate and less so for specific foods.
The factors were derived over a century ago and research has found they overestimate available energy from some foods, particularly whole nuts and high-fibre foods, since not all of it is absorbed.
Tolerances permitted in labelling are also wider than most people assume.
Processing changes availability
The same food processed differently delivers different amounts of energy.
Finely ground, cooked and softened foods are digested more completely than whole, raw, structurally intact ones.
Which means whole almonds and almond butter with identical panels do not deliver identical energy, and the panel cannot express this.
The matrix effect
Nutrients behave differently depending on what surrounds them.
Fat in cheese appears to affect blood lipids differently from the same fat in butter, which has been observed in several studies and is attributed to the food structure.
Which is one reason nutrient-based advice has moved toward whole-food and dietary-pattern advice.
Satiety
Not on the panel and among the most consequential properties.
Foods with identical energy content produce very different fullness, driven by protein content, fibre, volume, water and texture.
Which affects total intake far more than the calorie figure of any single item.
Glycaemic response
Carbohydrate content says nothing about how quickly it raises blood glucose.
Which depends on the food's structure, fibre, fat and protein content, cooking method and what else is eaten alongside.
Individual responses to identical foods also vary substantially, which has been demonstrated directly.
Ingredient quality
Two products with identical panels can have entirely different ingredient lists.
Which means the panel does not distinguish between a food made from whole ingredients and one assembled from refined components with the same nutritional totals.
The ingredient list carries that information and is generally more informative than the panel for that purpose.
Portion presentation
Per-portion figures depend on a portion size the manufacturer defined.
Which makes the per-hundred-gram column the only reliable basis for comparison between products.
What to actually do with a label
Compare per hundred grams between products in the same category.
Read the ingredient list, which tells you what the food actually is.
Treat calorie figures as approximate rather than precise.
And weight satiety and ingredient quality alongside the numbers, since neither appears on the panel.
Anyone managing a medical condition through diet should be working with a registered dietitian rather than interpreting labels alone.
Serving suggestions and preparation
Figures generally refer to the product as sold rather than as prepared.
Which means a dried product's figures do not include what is added to it, and a product requiring cooking in oil does not include the oil.
Where both are given, checking which column applies avoids substantial errors.
Reference intakes
Percentage figures relate to a reference adult with an assumed energy requirement.
Which is not a personal target, and requirements vary considerably with size, age, activity and sex.
Treating the reference as a personal allowance is the most common misuse of the panel.
Salt and sodium
Different jurisdictions declare one or the other, and they are not the same number.
Which means comparing products labelled differently requires a conversion, and the factor is published.
What is worth doing
Reading the ingredient list first, since it describes the food.
Using the panel for comparison within a category rather than as an absolute judgement.
And treating the whole thing as one input among several, since a food is not adequately described by seven numbers.
Front-of-pack schemes
Colour codes and grades summarise the panel into a simpler signal.
Which helps comparison at a glance and necessarily loses information, and the algorithms behind different schemes weight nutrients differently.
A product can therefore score well under one scheme and poorly under another, which is a genuine limitation of summarising.
What is not measured at all
Phytochemicals, most micronutrients, fibre types and the structure of the food are all absent.
Which covers a great deal of what current nutrition research considers relevant, and it is a limitation of the format rather than a failure of it.
Where the panel is most useful
Comparing similar products, checking salt content, and identifying products where added sugar is high relative to the category.
Which are real uses within the limits described, and they are what the format was designed for.