Nova’s food classification system warrants caution

Several recent epidemiological studies have suggested an association between ultra-processed food consumption and rates of obesity, cardiovascular disease or all-cause mortality, when applying a food classification tool known as Nova to their data.

Nova segregates foods and ingredients into four categories: minimally processed (eg fruits, grains, meat, milk); processed culinary ingredients (eg vegetable oil, salt, sugar); processed foods usually comprising two or three ingredients (eg canned vegetables or fruit, salted/sugared nuts, artisan bread); and ultra-processed foods and drinks with five or more ingredients and/or containing additives or particular ingredients (eg sweets, sugar-sweetened beverages, non-artisanal bread, breakfast cereals, tomato/vegetable- based pasta sauces).

Ultra-processed foods​

Thus, the latter category is a mixture of so-called discretionary foods that should not be the mainstay of healthy diets, alongside processed foods that can be important nutritionally for many people, if consumed in appropriate amounts.

A limitation of this type of study design is that it can’t determine a causal relationship. But, perhaps more importantly, Nova does not take into account the overall nutrient content of the foods or their presence in existing national dietary guidelines.

Categorising dietary fibre ​

So, important sources of dietary fibre for many – non-artisan wholemeal bread and high-fibre breakfast cereals – are categorised as ultra-processed. In a recent study, 24% of foods classified as A (of good nutritional quality) in the French Government’s NutriScore system were classed as ultra-processed by Nova.

Helping people get a handle on the nutritional quality of foods is clearly important in driving healthier habits, but it is debatable whether the Nova classification achieves this, with its focus on degree of processing per se, over nutritional value.