As the FSA began the process of reviewing its controversial Nutrient Profiling Model this week, it was on the receiving end of more criticism.

Dr Adam Drewnowski, a leading nutritionist and expert in obesity, published a peer-reviewed report in US scientific journal Nutrition Today damning the NPM as little more than a calorie-counting device - pointing out that 53% of the variance in a product is determined by the number of calories alone.

With pressure building in some circles in the US to consider a nutrient profiling model as a weapon against child obesity, Drewnowski's report dismissed the FSA model as "crude" and called for better science. "The NPM tells you little more than you can learn from an assessment of the amount of calories per 100g," says the director of the University of Washington's centre for obesity research. "It penalises foods that are dry, such as grains, because they are low in water, and does not consider many of their nutrients.

"The application of scores based on negative nutrients may lead to whole categories of food failing the nutrient profiling test. It's questionable how well it assesses cakes, cookies, cereals, or dry products because it focuses disproportionately on energy density."

Treating fruit, vegetables and nuts the same way regardless of their very different nutritional values is totally unscientific, he says. " You can't promote specific food groups and claim the results are based on science."

The report also challenges the highly subjective methods used by the 700 nutrition professionals asked to rate 120 foods in an attempt to validate the NPM. "The FSA asked nutritionists to assess foods against pre-determined definitions of healthy food, so of course there was a close link," he adds. "But some of the suppositions should be challenged. It needs other measures in place to ensure that the assessment is fair."

His criticism comes as the US looks at new ways to tackle its child obesity crisis, which is worse than anywhere else in the world.

Ominously, the FDA in the US is understood to favour a UK-style approach based on the NPM. Other ideas being considered are sticking with the current practice of emphasising positive nutrients in foods, or developing a third way.

"There is an internal debate in the US about whether we want to stand up against the UK approach, or create a score that incorporates those things but provides necessary balance and doesn't just look at sheer calories," he says.

Drewnowski has developed an alternative model, the Naturally Nutrient Rich score, which he believes irons out many of the NPM's problems because it takes into account 14 positive nutrients including vitamins, minerals and micronutrients, while the NPM only counts protein and fibre. "The NNR aims to deal with the absence of nutrients in a lot of US diets," he says. "We want to sufficiently focus on the positives."

Drewnowski believes there are "dangers inherent in basing a nutrient density score on too few or too many nutrients. A score based on a large number of vitamins, minerals and other micronutrients may have little discriminating power. Selecting the optimal number of nutrients is one priority for future research."

Another alternative the report explores is the development of different nutrition criteria for separate food groups. This is being considered by the French for their own nutrient profiling model, but it's difficult to come up with a definitive list of categories, warns Drewnowski. The Danes have 26 food groups while the Dutch have only 14. Plus, categories are often culture-specific - the French include a subcategory for ­croissants.

Cost must also be considered. Preliminary calculations using 2006 Seattle prices showed that of 375 foods tested, those that passed the NPM were on average four times more expensive than HFSS foods. "Nutrient profiling models should try to take cost into account," he says. " They ought to be sensitive to the average shopper."

The FSA refuted Drewnowski's findings. "We recommend that people have a balanced diet and this doesn't have to be expensive," says a spokeswoman. "The NPM is not about dietary advice."

Drewnowski has promised to continue to examine alternatives, however. In January he will publish a "big article" in Nutrition Reviews, the International Life Sciences Institute journal, detailing ways to develop nutrient profiles. A third paper, for the Journal of Nutrition, will examine further the close relationship between energy density and the FSA score.the NPM flaws

More than 50% of the variance in scores is determined by energy density (ie calories) so in many cases the same categorisation of food could have been achieved just by counting calories.

Dry products, such as grains, cereals and biscuits, and dried fruit such as raisins are unfairly penalised because of their low water content.

NPM fails to take into account essential vitamins and minerals and other important positive nutrients in products.

It only considers food and drink based on a 100g portion rather than the individual's recommended calorie intake.

The cost of products that pass the NPM are on average four times as expensive as those that are classified as HFSS.

The methodology of validating the NPM system using expert opinion is subjective because personal and cultural points of view will have also been used to assess foods.

Source: What's next for Nutrition Labeling and Health Claims?