Ref. Ares(2021)6170379 - 11/10/2021
Unilever requests for changes to Nutri-Score algorithm
Here are the key areas where the Nutri-Score (NS) algorithm should be changed:
1. Calculation per 100g
The current nutrient profile is based on per 100 grams, this results in the amount of nutrients
consumed being overestimated for small portions and underestimated for large portions. NS should
be adapted to be portion based (but we realize this is highly unlikely in the absence of EU regulated
serving sizes) or product group specific, as this is a better approach to:
• help consumers make the healthier food choice
• stimulate the foods industry to reformulate towards healthier products
• be based on evidence of impact on making better choices
We propose an algorithm based on a limited number of product groups that align to existing
schemes
(EFSA, EU Pledge, Choices International and
Keyhole). Appendix 1 shows more details on
the product groups but in summary:
• Four product groups (Meals, Beverages, Diary, Fats) are an adaptation of the current product
groups used by NS
• Additional five new groups (Fruit and vegetables, cereal/carbohydrates, meat/fish, sauces
and condiments, snacks and treats).
2. Acceptable Level of Processing for Fruit and Vegetables
The current guidance for NS state that powdered fruit and vegetables and freeze dried vegetables
and fruits cannot be included in the fruit and vegetable calculations. Published data from the
nutritional quality of dried soups that shows that the nutrient content is retained, and as such they
should be included in the calculation.
Ref: Leo van Buren et al. Nutritional Quality of Dry Vegetable Soups. Nutrients. 2019 Jun; 11(6): 1270.
3. Concentrated Fruit Juices
The differences between concentrated fruit juices when used as an ingredient and when used as the
basis for a fruit drink should be removed.
• Currently fruit juices made from 100% concentrates, e.g. orange juice made from
concentrate that is rehydrated to 100% can be counted towards fruit and vegetables.
However concentrated fruit juices or fruit puree that have not been rehydrated to 100% e.g.
juice/concentrated lemon syrup used in a sorbet cannot be counted as fruit/vegetables.
4. Rehydration Factor
The NS guidance propose a standard ‘rehydration factor of 2’ to rehydrate all dehydrated and
concentrated products. The consequence of this is that a double tomato concentrate, or a triple
tomato concentrate would have the same rehydration factor applied, even though they clearly do
not contain the same quantity of equivalent fresh tomatoes.
This is also in conflict of EU regulations which explicitly addresses the effect of dehydration and
dilution. In such regulation there are provisions which mandate that any concentration or dilution
factor shall be ‘provided and justified’ (see for example article 2 in Regulation (EC) No 1881/2006).
Furthermore such rehydration factor of 2 is also in conflict when any of these dried fruits or
vegetables would contribute to a nutrition or health claim in the final product as those claims shall
refer to the food ready for consumption (article 5-3 in Regulation (EC) No 1924/2006), so actual
rather than standard rehydration factors must be applied.
Even when dried fruit are used in a cereal bar or a cake, that are consumed without being
rehydrated a rehydration factor should be applied to calculate NS, which is illogical.
5. Plant Protein
Soya protein can gain points for protein but not fruit and vegetables in the NS calculation. However,
products that use more novel based plant-based proteins can gain points for both protein and fruit
and vegetables. This is approach is inconsistent as soyabeans are considered a legume and the soya
texturization process does not reduce its nutritional value but improves it.
6. Whole grain
In many countries, dietary guidance refers to whole grain products, yet the NS does not account for
this. The addition of clear rules on wholegrain ingredients would be welcomed.
7. Oils
Nutri-Score only includes rapeseed, walnut and olive oils in the ‘positive component’, indicating that
these are the only three oils that “reflect the public health recommendations that advocate to
favour these oils compared to other fats”. However, to align with public health messaging we would
like to see the addition of other unsaturated fats/oils as they too can be used in the diet to replace
fats that are high in saturated fat. This aligns to the recommendations of dietary guidelines in e.g.
Guidelines e.g. NL, UK, Denmark
8. Rounding Rules
The scoring of nutrients should be aligned with the rounding rules in the EU. For example:
• For zero calories / zero sugar beverages that declare <4kcal/100ml and <0.5g/100ml sugars
they score less favourable that if the nutrition information were declared as 0kcal/100ml
and 0g/100ml sugars, despite having the same meaning under EU Guidance (see table 4
therein).
General Comments
Here are some additional general comments:
Continue to be an all-inclusive model
The scheme should continue to be applicable to all products and not leave out any specific product
groups, to be as consistent as possible for the consumer. For example, excluding artisanal
unwrapped cheese but labelling packaged cheese.
Communication of the score
The FOP label should help encourage consumers to choose the better option within a product group,
however it should not prevent consumption of certain foods or food groups. This is particularly
challenging when there is not good within product group differentiation. Certain product groups
have a score in the range of A and B whereas others only score D and E.
Through communication campaigns this should be explained to the consumer, to promote healthier
choices, where they can be taught to look at a product group and choose D over E. However, we do
see the front of pack labelling scheme being used in applications such as retailer promotions and
marketing to children, where only A and B products are favoured.
Effectiveness study
In market effective studies, after implementation, to ensure good insight in how consumers use the
schemes to improve their diet is important. Only with these learnings can we evaluate the
effectiveness and adapt the scheme to maximise public health benefit.
APPENDIX 1
