Are Nutrition Facts Enough to Evaluate Ultra-Processed Foods?

“Processed” and “ultra-processed” are not synonyms

Washing, freezing, pasteurisation, fermentation, and canning are forms of processing that may improve safety, shelf life, and access. It is therefore inaccurate to assume that all processing is harmful.

Under NOVA, ultra-processed foods are industrial formulations often made largely from substances derived from foods and may contain ingredients uncommon in home kitchens or additives intended to make products convenient, durable, or highly appealing. Sugary drinks, packaged snacks, confectionery, and some ready meals are typical examples. Yet the category is broad and includes products with very different nutritional characteristics.

Why might the nutrition panel be insufficient?

Nutrient-profile systems usually focus on measurable amounts of energy, sugars, sodium, saturated fat, protein, and fibre. That is useful, but it may not capture disruption of the food matrix, eating rate, satiety, or the combined use of additives.

A product can improve its nutrient score by reducing sugar or adding fibre and protein while remaining a highly refined, rapidly consumed formulation. Conversely, a traditional food containing olive oil or salt may score less favourably on one nutrient axis without being ultra-processed. The classifications answer different questions.

What does the evidence show?

A large umbrella review of observational evidence associated greater ultra-processed food exposure with numerous adverse outcomes, particularly cardiometabolic outcomes, common mental disorders, and mortality. These findings establish associations, but observational studies alone cannot prove causation or fully remove socioeconomic, lifestyle, and measurement-related confounding.

Controlled feeding studies strengthen causal inference. Hall and colleagues found that an ultra-processed diet produced greater energy intake and short-term weight gain even when the presented nutrient composition was broadly matched. In a longer crossover trial published in 2025, both diets followed healthy dietary guidance, yet the minimally processed diet produced greater weight loss. Eating rate, energy density, texture, palatability, and food matrix may therefore matter beyond the nutrition panel.

Still, ultra-processed foods form a heterogeneous category. Current evidence does not justify assuming that every subgroup has the same effect or that every additive causes clinical harm in humans. Research now needs to separate which product characteristics contribute to which outcomes.

Why are additives part of the discussion?

Emulsifiers, sweeteners, colours, thickeners, and flavours are not one uniform biological category. Safety is assessed by substance and permitted use level. A long ingredient list is therefore not automatically dangerous, just as a short list is not automatically healthy.

Uncertainties remain around combinations, long-term exposure, and susceptible groups. Human trials of selected emulsifiers are exploring possible microbiome and metabolic effects, but these results cannot be generalised to all additives. The evidence should be neither dismissed nor overstated.

How can we decide in daily life?

Start with the whole product rather than one ingredient or score. Read the first ingredients, realistic portion size, and nutrition panel, then consider the product’s role in the diet. Plain yoghurt, frozen vegetables, canned legumes, and wholegrain bread are processed; that fact alone does not make them foods to avoid.

The goal is not a perfect shopping basket. Build most meals from vegetables, fruit, legumes, whole grains, nuts and seeds, suitable dairy foods, and quality protein sources. Instead of absolute bans, begin with ultra-processed products that are consumed most often and are easiest to replace.