As part of the Childhood Obesity Plan, Public Health England (PHE) was asked by the Department of Health to review the UK Nutrient Profiling Model to bring it in line with current UK dietary recommendations, in particular the changes to recommendations for fibre and free sugars. The review included developing a modified Nutrient Profiling Model, which was published for consultation on 24th March 2018.
The current Nutrient Profiling Model was developed in 2004-2005 by the Food Standards Agency. It has been used by Ofcom since 2007 to identify foods and drinks high in fat, sugar and/or salt, which then cannot be advertised during children’s television programmes and, since 2017, in non-broadcast media. This model covers foods and drinks on a 100g basis and uses a scoring system based on four ‘negative’ factors (energy, total sugars, saturated fat and sodium) versus three ‘beneficial’ factors (fruit, vegetables and nuts, fibre and protein). Scores related to each of these are calculated and the total score from the ‘beneficial’ factors subtracted from the total score of the ‘negative’ factors. A food or drink scoring above a certain threshold is classed as ‘less healthy’ and cannot be advertised. The thresholds are four for a food and two for a drink.
The 2004-5 nutrient profiling model was used as the starting point for the review, which began in 2016. The modelling was undertaken by Public Health England, and an expert group and reference group were established to review and support the process. Prof Judy Buttriss, Director General of the Foundation was a member of the expert group.
The key changes proposed to the 2004-5 nutrient profiling model are summarised below:
- The energy value has been revised slightly to 2000kcal (8.4MJ) to reflect the figure generally used for Government advice and on food labelling. Previously a higher value was used as an average was taken of older and younger children’s requirements.
- Sodium has been replaced with salt, using 6g as the reference value.
- Total sugars has been replaced with free sugars and the reference value reduced to reflect the change in recommendation for free sugars (formerly known as non-milk extrinsic sugars) from 10% to 5% of total dietary energy.
- Fibre values (now measured using the AOAC method rather than as non-starch polysaccharide (NSP)) have been updated to reflect the increased recommendation for dietary fibre of 30g per day, which is used as the reference value in the model.
The reference amounts for protein and for the sum of fruit, vegetables and nuts have remained the same.
These revisions are the result of the in-depth modelling exercise, where different scenarios were considered, primarily for fibre and for sugars, and their effect tested in terms of the foods and drinks that were passing and failing the model compared with the 2004-5 model (that is, falling above or below the threshold of four for foods and two for drinks) . It was agreed at the start of the work that the changes in government policy and recommendations (referred to above) with regard to free sugars would mean that fewer foods and drinks high in free sugars should pass the model. Salt, saturated fat and fibre were also monitored to check for unintended consequences in terms of foods and drinks passing the model that were inappropriate.
The modelling work was conducted using a test dataset of over 2000 food and drinks, for which nutrient values were assembled, including values for free sugars. The free sugars values all had to be calculated (as they did not previously exist in food composition databases). The details of the assumptions made for each of the food categories are outlined in appendix I of the report and a decision tree on the process of estimating the free sugars content of foods and drinks is provided in appendix L.
The substantial change in the recommendation for free sugars – falling from 10% of total energy intake to 5% - has meant that 16% fewer foods and drinks in the test dataset that are high in free sugars pass the proposed model (i.e. could be advertised to children) compared with when the 2004-5 model is applied. Although more points can be scored for fibre content in the revised model, to reflect the change in recommendations (and the increased emphasis on the importance of dietary fibre), the increase in fibre recommendation is relatively small compared with the change in the sugars recommendation. As a result, some fibre-containing foods that currently pass the 2004-5 nutrient profiling model fail the proposed revised model because, as well as fibre, they contain substantial amounts of free sugars.
The Nutrient Profiling Model 2018 review is now out for consultation until the 15th of June 2018.