Transparency and science crucial for nutrient profiling, says expert
be transparent, and be validated with respect to healthy diets,
says a new paper that sets out the steps to developing and
validating such models.
Nutrient profiling is defined as the science by which foods are raked according to their nutritional composition. While nutrient density was previously applied to overall diets, increasingly it is being applied to individual foodstuffs. Much attention has been given to the term in the last couple of years, partly since the new European nutrition and health claims regulation requires that only foods with favourable nutrient profiles should be allowed to make claims. The model for this piece of legislation has not yet been finalised, but it must take into account total fat, saturated fat, trans fat, sugar and salt - all of which are seen as disqualifying nutrients, too much of which could stand in the way of a claim. As well as underscoring legislation, nutrient profiling is also seen as a key to helping consumers make good diet choices (in guidelines like MyPyramid in the US), labelling schemes, and food marketing and advertising restrictions (particularly to children). As Dr Adam Drewnowski, director of the Nutritional Sciences Program and professor of epidemiology and medicine at the University of Washington in Seattle, USA, recognised in his editorial in Nutrition Reviews, there are now a number of different efforts underway to develop, validate and test nutrient profiling models. However, he said that any model needs to reconcile a number of competing objectives. For instance, for regulatory purposes an algorithm is needed that is "scientifically justifiable, enforceable, objective and transparent," he said. On the issue of transparency, he noted that "public health has no need for profiling models that are patented or proprietary". The author added that there have been some concerns from the food industry about nutrient profiling models making a judgement call as to whether a food is 'good' or 'bad', thereby penalising some categories. An example of this is the UK Food Standards Agency's traffic light labelling scheme. Drewnowski suggested that one way of combating this would be to create profiles that are category-specific, rather than across the board. "Despite such as diversity of goals, the development of nutrient profile models should follow uniform, rigorous and science-driven rules," he wrote. In his paper, Drewnowski set out recommendations he sees as important:
The index nutrients need to be relevant to the dietary needs and, preferably, limited in number;
The reference daily values should be based on an authoritative source and linked to food labelling;
The algorithm ought to be simple and transparent.
The models must be validated against independent measures of a healthy diet, and ideally against health outcomes.
The models need to be weight against food cost and enjoyment.
As a last word, Drewnowski said that consumer research is an integral part "…The ultimate test is whether nutrient profile models will communicate nutrition information in a way that is both useful and valuable to the consumer."
