Simplifying the complexities on the new NPM model – 2004 vs 2018 models

What is the Nutrient Profiling Model?

Think of the UK Nutrient Profiling Model (NPM) as a nutritional scorecard for food. It was created by the UK government to identify which foods are “less healthy”. The current NPM serves as the regulatory backbone for HFSS legislation, which by way, regulates the location of food products in retailer shelves, and restricts their advertisement permissions on TV and online.

The calculation

The result is gathered by subtracting the positive points – those from ingredients we should consume more of, like fruit, vegetables, nuts, seeds, fibre, and protein – from the negative – those from ingredients we should limit, like calories, saturated fat, salt, and sugar. If the final score is more or equal to 4, the food is classified as HFSS (High in Fat, Salt, or Sugar).

The HFSS Calculation: Negative Points – Positive Points = ?

Food products scoring ≥ 4 are HFSS
Drink products scoring ≥ 1 are HFSS

What are the differences between the 2004 and 2018 NPM?

On the 27th January 2026, the Department of Health and Social Outcome accepted the NPM 2018 updates of the 2004/05 framework to align with current dietary advice and the Eatwell Guide, including SACN guidance on free sugars, and acknowledging the benefits of seeds, and updating the fibre scoring.

We’ve summarised the key differences here:

NPM 2004/2005 Model	NPM 2018 Model
Sugar Threshold	Based on 21% of food energy from total sugars	Based on 5% of food energy from free sugars
Fibre Weighting	Max score of 5 points	Max score of 10 points (increased incentive for high-fibre foods)
Energy Reference	Based on 8,950 kJ (approx. 2,130 kcal)	Based on 8,400 kJ (approx. 2,000 kcal)
Positive Components (FVN/S)	Included Fruit, Vegetables, and Nuts	Included Fruit, Vegetables, Nuts, *and Seeds*

Reference: BNF, 2026; Department of Health & Social Care and Public Health England, 2026

Why are updates to the NPM important?

The 2018 model uses the free sugars definition recommended by the Scientific Advisory Committee on Nutrition (SACN) and the World Health Organization (WHO), so that it focuses on how the body and teeth process the sugar. When you blend fruit into a smoothie, purée, or syrup, you break the cell walls, “freeing” the sugar. This sugar can have a higher cariogenic effect on your teeth and have detrimental consequences on your blood sugars. Using free sugar values rather than total sugar values means that sugars in intact fruit and vegetables will not be considered as ‘negative’.

The raising of the fibre weighting from five to ten means that foods high in fibre may now accrue more positive points, incentivising their introduction during reformulation. Yet, on this note, not all fibres can count towards fibre points, as expressed by the SACN report Carbohydrates and Health (2015). Those that are synthetic and are not demonstrated by reliable scientific evidence to provide beneficial physiological effects cannot be counted as fibre. Similarly, ingredients added to boost fibre that cannot be measured by the standard AOAC lab test, do not count towards fibre points.

Finally, the inclusion of seeds as “positive”/C points is aimed for manufacturers to benefit from the addition of seeds into their products. See the following sources to view the 2004 vs 2018 model A and C points table:

The challenges retailers are facing with these changes

1. The “Free Sugars” Calculation Barrier

This is widely considered the biggest technical hurdle. Unlike total sugars values, free sugars are not required to be labelled ‘on pack’, Determining free sugars requires the ingredient amounts as well as the quantities of free sugars within these ingredients – this data is often not available to retailers.

2. The “FVNS Points” Calculation Barrier

Another similar challenge is calculating the quantities of ‘positive’ components in products that count towards the FVNS points. Again, without the recipe data – it is difficult to score products accurately, and at scale. Retailers often have to rely on suppliers to provide this information, and if it is not available, often erring on the side of caution, and giving a 0 for FVN points is a method often used.

Spoon Guru’s Solution

Spoon Guru’s technology is able to overcome the challenges of estimating the free sugars and FVNS content of products, applying the NPM 2018 methodology consistently and at scale.

Here’s how Spoon Guru’s technology handles the NPM 2018 methodology:

Smart Categorisation:

Our Natural Language Processing AI automatically sorts ingredients into healthy groups (Fruits, Veg, Nuts, Seeds).

➡️ Our Natural Language Processing (NLP) technology enables us to categorise ingredients in products into their relevant FVNS groups in detail. This includes applying multiplication factors for concentrated tomato purees and dried fruits and veg.

Automated Estimates:

Our FVNS algorithm calculates ingredient percentages even when they aren’t explicitly listed, ensuring precise nutritional scoring.

➡️ Our FVNS algorithm identifies QUID (Quantitative Ingredient Declaration) percentages from the categorised ingredients and estimates the quantities of these on a ranking basis where QUID is not available, allowing FVNS points estimation.

Free Sugar Mapping:

We link ingredients to data models that accurately estimate hidden free sugars.

➡️ Categorised ingredients are mapped to published free sugars estimates to enable free sugars content estimation of each ingredient.

Process-Aware:

We factor in how food is processed, strictly following the latest UK Department of Health guidelines.

➡️ When calculating the total free sugar amount, a product’s degree of processing is also taken into account, following the technical guidance by the British Department of Health and Social Care on the Nutrient Profiling Model 2018 (27 January 2026)

Further developments such as considerations regarding the added fibres – i.e. extracted natural carbohydrate components or synthetic carbohydrate products that do not always count towards the positive points, identifying the added sugars from lactose-free products, and validation of the accuracy of our methodology, are under current development.