The Thumbs food classification system was developed to assist remote Australian communities to identify food healthiness. This study aimed to assess: (1) the Thumbs system’s alignment to two other food classification systems, the Health Star Rating (HSR) and the Northern Territory School Canteens Guidelines (NTSCG); (2) its accuracy in classifying ‘unhealthy’ (contributing to discretionary energy and added sugars) and ‘healthy’ products against HSR and NTSCG; (3) areas for optimisation. Food and beverage products sold between 05/2018 and 05/2019 in fifty-one remote stores were classified in each system. System alignment was assessed by cross-tabulating percentages of products, discretionary energy and added sugars sold assigned to the same healthiness levels across the systems. The system/s capturing the highest percentage of discretionary energy and added sugars sold in ‘unhealthy’ products and the lowest levels in ‘healthy’ products were considered the best performing. Cohen’s κ was used to assess agreement between the Thumbs system and the NTSCG for classifying products as healthy. The Thumbs system classified product healthiness in line with the HSR and NTSCG, with Cohen’s κ showing moderate agreement between the Thumbs system and the NTSCG (κ = 0·60). The Thumbs system captured the most discretionary energy sold (92·2 %) and added sugar sold (90·6 %) in unhealthy products and the least discretionary energy sold (0 %) in healthy products. Modifications to optimise the Thumbs system include aligning several food categories to the NTSCG criteria and addressing core/discretionary classification discrepancies of fruit juice/drinks. The Thumbs system offers a classification algorithm that could strengthen the HSR system.

In the context of the global childhood obesity, it is essential to monitor the nutrition value of commercial foods. A cross-sectional study (November 2018 to April 2019) aimed to evaluate the nutritional adequacy of processed/ultra-processed food products targeted at 0–36-month-old children in Portugal and in Brazil. The nutrient profiling model developed by the Pan American Health Organization was used. A total of food 171 products were assessed (123 in Portugal and forty eight in Brazil). From the fifteen available meat- or fish-based meals in Brazil, 60 % exceeded the amount of Na and 100 % exceeded the target for total fat. Given the lack of specification of sugars within carbohydrates in the label of the foods in Brazil, it was not possible to calculate free sugars. In Portugal, from the seventeen fruit and vegetable purees and the six juice/smoothie/tea/drinks available, 82 % and 67 %, respectively, surpassed the level of free sugar, while total and saturated fat was excessive in all yogurt and yogurt-related products (n 21), 40 % of biscuit/wafer/crisps (two out of five) and 13 % meat- or fish-based meals (two out of sixteen). These findings demonstrate the relevance of improving the nutritional profile of some food products targeted to young children.

The Nestlé Nutritional Profiling System (NNPS) has been developed to guide food and beverage reformulation. The WHO published guidelines to develop and validate nutrient profiling systems. The objective was to conduct validation tests of the NNPS following principles of the WHO guidelines. French (Individual and National Survey on food Consumption 2006–2007) and the USA (National Health and Nutrition Examination Surveys 2011–2012) nationally representative dietary surveys were used. NNPS outcomes (PASS, FAIL, out-of-scope) of foods were compared with the validated UK Ofcom nutrient profiling system outcomes. Contributions of NNPS outcomes to energy intakes were compared between diets nutritional quality classes defined by two methods: based on a food-based quality indicator (Programme National Nutrition Santé Guideline Score in France, Healthy Eating Index 2010 in the USA) or on a combination of three nutrient-based indicators (mean adequacy ratio, mean excess ratio and energy density). In both countries, food items with a NNPS FAIL outcome had a lower nutritional quality according to the UK Ofcom, with an overall agreement between the two systems of 75·7 % in France and 68·8 % in the USA. In both countries, a high (respectively, low) contribution of NNPS PASS (respectively, NNPS FAIL) was positively associated with diet healthiness. Absolute associations were stronger between the contribution of NNPS FAIL products and measures of diet healthiness. Foods and beverages reaching NNPS standards appeared to have a higher nutritional quality and would be more likely to contribute to healthier diets, mainly linked to a reduction of nutrients to limit.

Nutrient profiling (NP) is a method for evaluating the healthfulness of foods. Although many NP models exist, most have not been validated. This study aimed to examine the content and construct/convergent validity of five models from different regions: Australia/New Zealand (FSANZ), France (Nutri-Score), Canada (HCST), Europe (EURO) and Americas (PAHO). Using data from the 2013 UofT Food Label Information Program (n15342 foods/beverages), construct/convergent validity was assessed by comparing the classifications of foods determined by each model to a previously validated model, which served as the reference (Ofcom). The parameters assessed included associations (Cochran–Armitage trend test), agreement (κ statistic) and discordant classifications (McNemar’s test). Analyses were conducted across all foods and by food category. On the basis of the nutrients/components considered by each model, all models exhibited moderate content validity. Although positive associations were observed between each model and Ofcom (all Ptrend<0·001), agreement with Ofcom was ‘near perfect’ for FSANZ (κ=0·89) and Nutri-Score (κ=0·83), ‘moderate’ for EURO (κ=0·54) and ‘fair’ for PAHO (κ=0·28) and HCST (κ=0·26). There were discordant classifications with Ofcom for 5·3 % (FSANZ), 8·3 % (Nutri-Score), 22·0 % (EURO), 33·4 % (PAHO) and 37·0 % (HCST) of foods (all P<0·001). Construct/convergent validity was confirmed between FSANZ and Nutri-Score v. Ofcom, and to a lesser extent between EURO v. Ofcom. Numerous incongruencies with Ofcom were identified for HCST and PAHO, which highlights the importance of examining classifications across food categories, the level at which differences between models become apparent. These results may be informative for regulators seeking to adapt and validate existing models for use in country-specific applications.

Obesity rates are increasing worldwide. Potential reasons include excessive consumption of sugary beverages and energy-dense foods instead of more nutrient-rich options. On a per kJ basis, energy-dense grains, added sugars and fats cost less, whereas lean meats, seafood, leafy greens and whole fruit generally cost more. Given that consumer food choices are often driven by price, the observed social inequities in diet quality and health can be explained, in part, by nutrition economics. Achieving a nutrient-rich diet at an affordable cost has become progressively more difficult within the constraints of global food supply. However, given the necessary metrics and educational tools, it may be possible to eat better for less. New metrics of nutrient density help consumers identify foods, processed and unprocessed, that are nutrient-rich, affordable and appealing. Affordability metrics, created by adding food prices to food composition data, permit calculations of both kJ and nutrients per penny, allowing for new studies on the economic drivers of food choice. Merging dietary intake data with local or national food prices permits the estimation of individual-level diet costs. New metrics of nutrient balance can help identify those food patterns that provide optimal nutritional value. Behavioural factors, including cooking at home, have been associated with nutrition resilience, defined as healthier diets at lower cost. Studies of the energy and nutrient costs of the global food supply and diverse food patterns will permit a better understanding of the socioeconomic determinants of health. Dietary advice ought to be accompanied by economic feasibility studies.

The primary goal of front of pack (FOP) labelling is to help consumers make healthier choices through communication. A secondary goal is to encourage producers to improve the nutritional composition of their products. Evidence has shown that (FOP) labelling can help consumers to make healthier food choices and has been an incentive for producers to improve product composition. As FOP labelling is seen as an important tool to improve food environments for public health purposes, the WHO supports initiatives of governments to implement an FOP labelling system. Based on the experiences of a wide range of countries over many years, possible success factors for such an FOP system have been defined, six of which are discussed in the present paper and used to evaluate the Dutch Choices Programme that was started in 2006. In the course of time a large number of producers joined the programme and the logo was recognised by more than 90 % of the consumers, but by 2016 the Dutch consumer organisation argued on the basis of their own research that a quarter of the consumers did not understand the colour coding of the logo and as a result the Dutch government decided to no longer support this logo and to introduce a nutrition app. The challenge that remains is to find a system that consumers understand well and that still encourages manufacturers of food to improve product composition. New technology-based data collecting initiatives might provide the right tools to develop such a system.

Nutrient profiling (NP) models rate the nutritional quality of individual foods, based on their nutrient composition. Their goal is to identify nutrient-rich foods, generally defined as those that contain more nutrients than calories and are low in fat, sugar and salt. NP models have provided the scientific basis for evaluating nutrition and health claims and regulating marketing and advertising to children. The food industry has used NP methods to reformulate product portfolios. To help define what we mean by healthy foods, NP models need to be based on published nutrition standards, mandated serving sizes and open-source nutrient composition databases. Specifically, the development and testing of NP models for public health should follow the seven decision steps outlined by the European Food Safety Authority. Consistent with this scheme, the nutrient-rich food (NRF) family of indices was based on a variable number of qualifying nutrients (from six to fifteen) and on three disqualifying nutrients (saturated fat, added sugar, sodium). The selection of nutrients and daily reference amounts followed nutrient standards for the USA. The base of calculation was 418·4 kJ (100 kcal), in preference to 100 g, or serving sizes. The NRF algorithms, based on unweighted sums of percent daily values, subtracted negative (LIM) from positive (NRn) subscores (NRn – LIM). NRF model performance was tested with respect to energy density and independent measures of a healthy diet. Whereas past uses of NP modelling have been regulatory or educational, voluntary product reformulation by the food industry may have most impact on public health.

In this paper, I first provide definitions of nutrient profiling and of a nutrient profile model. I set out the purposes of nutrient profiling: both general and specific. I give two examples of nutrient profile models that have been developed for regulatory purposes by the Food Standards Agency (FSA) in the UK and the WHO for its European Region – the UK FSA/Ofcom and the WHO-Euro models – and compare the way the models are constructed and function, how they have been developed, the extent to which they have been tested and validated and their use in regulation. Finally I draw some conclusions about the future use of nutrient profiling for regulatory purposes. I argue that its full potential has yet to be realised and give some reasons why. I pose some urgent research questions with respect to nutrient profiling.

The food industry holds great potential for driving consumers to adopt healthy food choices as (re)formulation of foods can improve the nutritional quality of these foods. Reformulation has been identified as a cost-effective intervention in addressing non-communicable diseases as it does not require significant alterations of consumer behaviour and dietary habits. Nutrient profiling (NP), the science of categorizing foods based on their nutrient composition, has emerged as an essential tool and is implemented through many different profiling systems to guide reformulation and other nutrition policies. NP systems should be adapted to their specific purposes as it is not possible to design one system that can equally address all policies and purposes, e.g. reformulation and labelling. The present paper discusses some of the key principles and specificities that underlie a NP system designed for reformulation with the example of the Nestlé nutritional profiling system. Furthermore, the impact of reformulation at the level of the food product, dietary intakes and public health are reviewed. Several studies showed that food and beverage reformulation, guided by a NP system, may be effective in improving population nutritional intakes and thereby its health status. In order to achieve its maximum potential and modify the food environment in a beneficial manner, reformulation should be implemented by the entire food sector. Multi-stakeholder partnerships including governments, food industry, retailers and consumer associations that will state concrete time-bound objectives accompanied by an independent monitoring system are the potential solution.

The WHO has called for governments to improve children’s food environment by implementing restrictions on the marketing of ‘unhealthy’ foods to children. Nutrient profiling (NP) models are used to define ‘unhealthy’ foods and support child-directed food marketing regulations. The aim of the present study was to assess the suitability of the South African NP model (SANPM), developed and validated for health claim regulations, for child-directed food marketing regulations. The SANPM was compared with four NP models specifically developed for such regulations. A representative list of 197 foods was compiled by including all foods advertised on South African free-to-air television channels in 2014 and foods commonly consumed by South African children. The nutritional information of the foods was sourced from food packaging, company websites and a food composition table. Each individual food was classified by each of the five NP models. The percentage of foods that would be allowed according to the different NP models ranged from 6 to 45 %; the models also varied considerably with regard to the type of foods allowed for marketing to children. The majority of the pairwise comparisons between the NP models yielded κ statistics >0·4, indicating a moderate agreement between the models. An almost perfect pairwise agreement (κ=0·948) existed between the SANPM and the UK Food Standards Agency model (United Kingdom Office of Communication nutrient profiling model), a model extensively tested and validated for such regulations. The SANPM is considered appropriate for child-directed food marketing regulations in South Africa.