Nutrient Profiling- fine line between good and bad

Feelings are running high when it comes to Nutrient profiling. Categorising foods according to their nutritional composition provokes reactions from food companies’ lobbyists such as ‘demonising’ or ‘blacklisting’ measures. There is fear of losing business, and there is no consensus on what model is most suitable to safeguard consumers with healthy options.
Elsevier Food International, Vol. 9, Number 2, May 2006
Lucien Joppen

The term ‘nutrient prTofiling’ (NP) was coined by the EU several years ago in the context of its ‘Nutrition and Health Claims Proposal’. A controversial issue, as NP seeks to answer the question if products with a relatively high amount of fat, sugar and sodium, would be able to carry certain health and nutrition content claims. This to avoid situations in which, for example, sugary products would be advertised as being low in fat or carrying some sort of health claim. The problem is: where do you draw the line? When does a product qualify for a certain health claim or not? The British Nutrition Foundation (BNF) called it "an immense challenge, almost impossible, to devise a viable, robust and scientifically supportable model," so from the outset different parties have different opinions while the stakes are high.

Furious protests
Especially scoring models -like the recent model of the British Food Standards Agency (FSA)
which intends to regulate advertising - have led to furious protests from the industry. Strangely enough, however, the industry was consulted when the FSA designed the model.  Christine Welberry, spokeswoman for the British Food and Drink Federation (FDF) explains: "We have had the chance to say our say, but in the end we are not pleased with the final model. Our main objective is that the starting point of the model has been highly subjective and lacks a scientific basis. First, two groups of ‘healthy’ and ‘unhealthy’ foods have been made. Then a model has been designed which classifies these products accordingly. It sort of becomes a self-fulfilling prophecy." Dairy UK, the ‘voice of the British dairy industry’, also points at anomalies in the model’s composition. "We question the omission of protein quality, and micronutrients and minerals such as calcium, phosphorus, potassium and magnesium, as well as the vitamins A and B", said Ed Komorowski, Dairy UK’s technical director. The dairy industry’s concern is that some of its products might be excluded from marketing to children. The association argues that some of the nutrients mentioned could make up for - for example - the fat content of certain dairy products.  Another point of criticism on the FSA model is that it judges foods per 100 grams and not per serving. This leads to other anomalies such as olive oil scoring higher (i.e. being less healthy) than a range of potato crisps or cakes, despite olive oil’s relatively healthy reputation.
"We have tested hundreds of models over the last two years and came up with a model that reflects the dietary guidelines of the SACN (Scientific Advisory Committee on Nutrition), which was involved in the model making process," said Mike Rayner, one of the makers of the FSA model.  According to him, most anomalies have been ironed out during refining of the model. He admits that the FSA’s choice for a scoring model is more or less politically fuelled. "The model aims to categorise products of which an aboveaverage consumption does not contribute to a healthy lifestyle. In the end, British children should consume less of certain products or ideally switch more often to other categories, for example, by eating an apple instead of a bag of potato crisps or a snack bar."
In contrast with the scoring model, the threshold model seems to have raised less controversy. One reason is that most threshold models do not classify foods across-the-board in terms of less healthy, intermediate or healthy. For industry lobbyists this is a comforting thought, as ‘their’ food or drink stuffs are not ‘demonised’, quoting the word used by certain spin doctors in the industry. Threshold models merely set minimum and maximum levels of respectively‘good’ or ‘bad’ nutrients. If products stay within certain levels, then they qualify for a positive health mark (e.g. the Swedish Keyhole symbol). But there are no implications if they do not qualify. Contrary to the FSA model, labelling schemes based on threshold models are voluntary.

Devising an industry-wide scheme
Regardless of all the pros and cons of various NP models, there are nutritional experts who say nutrient profiling does not serve any purpose from a health educational point of view. "There are a number of issues which undermine nutrient profiling," says André Huyghebaert, former professor with the Department of Technology and Nutrition at the University of Ghent (Belgium). "The difference between composition of a food and actual consumption is very important. NP models do not take into account the amount, which means that the impact of a low level of a certain nutrient in great quantities is higher than a high level in limited quantities. Furthermore, NP does not cover all eating opportunities. Consumers mostly eat meals and not individual food items. The increasing out-of-home consumption also minimises the effectiveness of NP."
In spite of Huyghebaert’s arguments, NP is there to stay. Of course, the concept does not necessarily have to be used in conjunction with labelling schemes or advertising regulation. Unilever uses an across-the-board threshold model as the backbone of its Nutrient Enhancement Programme, designed to bring down the saturated and trans fat, sodium and sugar content of its portfolio.
Gert Meijer, vice president nutrition and health at Unilever: "We would prefer an industry-wide approach, not a companyspecific initiative. We have looked at national labelling schemes such as the Swedish Keyhole. However, these programmes are mostly confined to certain countries and the prevention of certain diseases. At the moment, we are talking with some of our competitors in order to devise an industrywide scheme."
Such a scheme would undoubtedly influence the impact of NP. If there were one labelling scheme, the impact on consumer behaviour could be higher. There is only one ‘but’. Which NP model should be used?