Anders Mller

1
Can we trust industry data?

• Anders Møller
• Danish Food Information, Denmark
• am_at_danfood.info

2nd EuroFIR Compiler Network meeting, Norwich,
March 2008
2
Two Important Reports

• Danish studyMikkelsen et al.Næringsdeklaration
  er og tolerancer(Nutrition Labelling and
  tolerances in Danish)
• European CommissionThe Introduction Of
  Mandatory Nutrition Labelling In The European
  UnionImpact Assessment Undertaken For DG Sanco,
  European CommissionService contract nr
  SANCO/2004/D4/EAS/SI2.378734

3
Danish Study (DFVA/DTU)

• Purpose
• The purpose of the survey Nutrition labelling
  and tole-
• rances has been to examine the nutrition label
  and the
• content of the declared nutrients for a number of
  prepack-
• aged foods, in order to examine to what extent
  the Danish
• guidelines for tolerances in nutrition
  declaration are met.
• The survey has had specific focus on certain
  foods which,
• due to their composition are known to be
  difficult to decla-
• re within the given tolerances.

4
Danish Official Tolerances
5
Study Design

• Part 1Study of 113 samples of ready made foods,
  frozen potato products, pies, pizzas, soups,
  charcuterie products, fish products and
  mayonnaise products were examined.Samples were
  chosen to examine the energy giving nutrients
  (protein, carbohydrate and fat), and dietary
  fibre and sodium for those products, for which
  the extended nutrient declara-tion was used.
• Part 252 samples of 13 different charcuterie
  products in small sized packages were
  examined.The purpose of this part of the study
  was mainly to examine the individual variation
  between packages of the same product in order to
  see whether the declared nutrients in each
  individual package were within the tolerance
  limits.

6
Main results I

• Part 1The survey showed, that in nearly half of
  the samples there was found a deviation from the
  nutrient declara-tion greater than the guideline
  tolerances, either for fat, carbohydrate or
  protein.
• In about 15 of the samples, there was more
  than one nutrient in the nutrition declaration,
  which differed from the declared values.However,
  there was a great difference between the product
  groups involved. The greatest deviations from
  tolerance limits were found for fish products,
  followed by ready made foods and charcuterie
  products. For soups and potato products no
  deviations were found.
• Comparing the declared nutrients in total, more
  deviations were found for fat than for
  carbohydrate and protein. Furthermore, there were
  found slightly more underestimated declared
  values than overestimated ones.

7
Main results II

• Part 2 The main purpose was to examine the
  variation be-tween individual small packages of
  the same product (charcuterie products). For
  rullepølse and beef brisket a relatively high
  number of deviations from the tole- rance limits
  were found (8 of a total of 9 products).
  Especially for rullepølse, which is a very
  inhomoge-neous product, a great variation was
  found between packages of the same product. For
  salami, which is a more homogeneous product,
  there was no significant difference between
  packages.

8

• ConclusionThe extent of deviations was modest,
  considering that the chosen samples were expected
  to have difficulties with respect to declaring
  the correct values, due to a great variation in
  the nutrient content of the raw material, and
  also due to an inhomogeneous distribution of
  ingredients in the products.