Consumer food products in some countries have a nutritional score, indicating how healthy the product is to eat. Each country has its own approach, but in the European Union the Nutri-Score in several countries.

This gem implements the Nutri-Score for Ruby and includes adaptations for different countries. The maintainer would be open to integrating other nutritional scores as well.

Note: this is currently under development, the API may change without notice, and scoring has not yet been fully verified. Be very careful using this in production.

There are currently two versions of the Nutri-Score in use. The first was developed in the UK by the Food Standards Agency and currently maintainted by the Department of Health. It is used for regulating food advertisements.

In France, an adapted Nutri-Score was adopted in 2017 for use as a voluntary label on the packaging. In 2018, Belgium and Spain adopted the same scheme.

Other countries are evaluating adopting the Nutri-Score as well, including The Netherlands. There was also a European citizen's initiative to adopt it Europe-wide, but this has been withdrawn.

gem install nutriscore

The input for all nutritional scores are a product category and nutritional values. Which nutrients are required depends on the product category. Note that fvnp stands for the amount of fruits, vegetables, nuts and pulses (and, since Oct 2019, also rapeseed, walnuts and olive oils).

require 'nutriscore'

# Fruit fromage frais
product_a = {
  energy: 459,          # kJ/100g
  fat_saturated: 1.8,   #  g/100g
  sugar: 13.4,          #  g/100g
  sodium: 0.1,          #  g/100g
  fvnp: 8,              #  g/100g (= weight-%)
  fibres: 0.6,          #  g/100g
  proteins: 6.5,        #  g/100g
}

# Compute the french Nutri-Score for a generic product.
score = Nutriscore::FR::SpecificScore.new(product_a)
#<Nutriscore::FR::SpecificScore score=0
#  positive_score=#<Nutriscore::FR::PositiveScore score=4
#                    fvnp=0 fibres=0 proteins=4>
#  negative_score=#<Nutriscore::FR::NegativeScore score=4
#                    energy=1 fat_saturated=1 sugar=2 sodium=0>>
score.score.single
# => 0
score.score_class.single
# => "B"

To be able to work with incomplete information, results are returned as ranges. The use of .single in the above example converts these to a single value (it returns nil if there is not enough information to get a single result). The following example shows what happens when data is missing.

score = Nutriscore::FR::SpecificScore.new(product_a.merge({ sodium: nil }))
score.score
# => 0..10
score.score.single
# => nil
score.score_class
# => "B".."C"
score.score_class.single
# => nil

Please only use #single and the regular Ruby Range methods on #score and #score_class. Other methods do exist, but are not guaranteed to be stable across releases.

Different categories can use different score classes:

• Nutriscore::FR::CheeseScore for cheese
• Nutriscore::FR::FatsScore for vegetable and animal oils and fats
• Nutriscore::FR::MineralWaterScore for mineral water
• Nutriscore::FR::DrinksScore for other drinks
• Nutriscore::FR::SpecificScore for other food products

Note that the score classes may use different nutrients (there is one additional nutrient, fat_total, used by FatsScore).

The UK has the same basis for computation, but it is used to determine whether a product can be advertised (it must not be less healthy).

score = Nutriscore::UK::SpecificScore.new(product_a)
score.score
# => 0
score.less_healthy?
# => false

The method #less_healthy? is UK-specific, and returns true, false, or nil if there is not enough information to make a judgement.

By default, the fibres measurement method is AOAC (which is preferred), but it is possible to use fibres values measured with the NSP method:

# Acceptable values for the fibres_method are: :aoac and :nsp.
score = Nutriscore::UK::SpecificScore.new(product_a, fibres_method: :nsp)

Different categories can use different score classes:

• Nutriscore::UK::SpecificScore for food products
• Nutriscore::UK::DrinksScore for drinks

This software is distributed under the MIT license.