equals
I could, perhaps, have talked more about how they weighted their choices so that certain nutrients were more important than others in their calculations - bottlenecks both the favorable and unfavorable, outliers, ect. I don't remember if these researchers do but some other researchers/writiers of this concept have weighted cholesterol particularly heavily because it is considered such a problem in a typical diet
Link to Reply to Food Compass novelty
"...Second, the Food Compass integrates cutting-edge science in selection of attributes,
including the use of nutrient ratios (unsaturated:saturated fat,
fibre:carbohydrate and potassium:sodium) which more accurately
incorporate overall fat, carbohydrate and mineral quality and
the biologic interactions between these components. With these
ratios, the Food Compass is the first major NPS to emphasize the
health benefits of foods relatively higher in unsaturated fats and
the harms of foods relatively higher in refined starches. The omis-
sion of attributes that are prominent in other NPS, in particular
total fat and total energy, is also a key advance that better reflects
modern nutrition science.
Third, the Food Compass scores foods on the basis of 100 kcal,
rather than per 100 g, as in many other NPS including NutriScore
and Health Star Rating. This is critical, as weight-based metrics can
be heavily biased by water content, as well as fibre and fat content.
For example, a recently published NPS concluded that pork fat was
the seventh most nutritious food in the world1. Upon review, these
conclusions are heavily driven by the bias of scoring per weight. For
instance, a 100 g portion of pork fat contains 638 kcal, while a 100 g
portion of apple contains only 52 kcal (due to high water weight).
Comparing the absolute amounts of nutrients provided by the same
weight of these two foods is a highly flawed comparison, given
the nearly 13-fold higher calories contributed by one food versus
another. Yet, many NPS score per 100 g. Food Compass scores per
100 kcal, a much more natural and unbiased unit of consumption...."
Peris, you will like this from the same paper (I do too),
"...the Food Compass uses nine domains, a
feature that provides a more holistic overview of healthfulness,
while also preventing excess influence of any single attribute—
and minimizing the ability of industry to manipulate or ‘game’
the score by adding fortified vitamins and minerals, which can
dominate the scores of other [Nutrient Profiling System]..."
Like all models, it works well for some aspects; not so much for others.
. This is really good of you to offer, we’ve got a local library here though so maybe that’ll be easier with the big screen.
