Can I modify the EPIC-Norfolk FFQ?
The FFQ can be modified but must be validated after modification. The program has been designed to enable new foods and nutrient data to be easily added to it. The FETA modification instructions explain how to do this.
What do I do if an FFQ has a lot of missing ticks in the main grid?
We recommend that FFQs having 10 or more missing ticks in the main grid are excluded since this level of missing data would lead to a considerable underestimate of intake.
Does the program use information from all questions of the FFQ?
No. Only questions 1, 3, 4, 5, 6, 7 and 10 are currently used by the program as only these have a direct effect on the nutrient intake.
Can I get information on intake of vitamin and mineral supplements?
This version of the FFQ does not provide intake data on supplements.
Is there a version of FETA that will run on a Mac computer?
We do not currently have a version of FETA that is compatible with Mac computers.
What are the differences between FETA versus CAFÉ processing?
In contrast to CAFÉ, which was only able to deal with a maximum of two breakfast cereals, FETA allows up to four cereal types to be recorded and portion weights are adjusted accordingly. Of the 24,633 participants who consumed breakfast cereal in the EPIC-Norfolk 1HC FFQ, 91% consumed either one or two cereals whereas 9% consumed either three or four breakfast cereals. The average portion size for all breakfast cereals is 30g, with the exception of muesli, which is 60g.
Some participants recorded porridge as one of their breakfast cereals. However, porridge consumption should be quantified in line 23 in Part 1 of the FFQ. In the CAFÉ program, porridge recorded in Part 2 was processed, but the FETA program excludes any food codes relating to ‘porridge’ type items from the cereal look-up list.
Frying and baking fats:
Occasionally, the free text entered for vegetable oil in Question 6 is the same as that entered for margarine in Question 7. In the CAFÉ system, only one food code could be assigned to unique free text. Therefore, if it was decided that ‘sunflower’ meant sunflower oil, (Question 6), when ‘sunflower’ was noted for the type of margarine used in baking, (Question 7), it would also have the food code of sunflower oil assigned to it. In the spreadsheet for FETA entry, improved layout and data entry ensures that the most appropriate food code is assigned.
Selection of the ‘none’ or ‘No’ box and default milk, cereal, and fat codes:
Sometimes, the ‘None’ or ‘No’ box in part 2 is ticked but further information provided and/or assumptions made, result in the assigning of a default code. For example, if an individual records that they consume milk in Question 4, but tick the ‘None’ box in Question 3, a default milk code is assigned.
A number of EPIC-Norfolk participants ticked the ‘none’ box for baking (N=3 925) and frying (N=903) fats, though it was thought more likely that fat was used but the type was unknown. In FETA, if the ‘none’ box is ticked, the appropriate default fat is used in the absence of free text, while in the CAFÉ program, ‘none’ was assumed to mean no fat used. The default baking and frying fats are taken from the Miscellaneous Foods supplement (19); these food codes used in the CAF#201; program have been changed in FETA to more appropriate codes. The default milk was calculated using 50% semi-skimmed milk, 25% whole milk and 25% skimmed milk (the default mapping used in CAFÉ was 50 % whole milk, 40% semi-skimmed milk and 10% skimmed milk).
These default codes are also applied by the program, as required, when specific food codes can not be assigned by text matching.