EPIC-Norfolk Food Frequency Questionnaire (FFQ) Entry and Processing Documentation


Welcome to our new FFQ processing tool, FETA (Φετα) FFQ EPIC Tool for Analysis, which is based on the earlier CAFÉ system (see the CAFE paper). This tool is free to use and will produce different levels of nutrient data, as well as basic food groups.

This program is based on version 6 (CAMB/PQ/6/1205) of the EPIC-Norfolk FFQ. The EPIC-Norfolk FFQ was originally developed in 1988 and its food list and portion sizes represent those of an adult population likely to have established eating habits and follow a traditional UK diet.

This questionnaire was validated for the EPIC-Norfolk population and its use may not therefore be appropriate for a different population group or a different research question. Even if used as the basis for developing another the FFQ, any changes to it would need further validation for the particular study population.



Need the EPIC-Norfolk FFQ for your study?

Please click here if you wish to download the FFQ to use in your study download the FFQ.


DATA PREPARATION

Entry of FFQ, including coding of free text

Your raw FFQ data can be coded either using the Access data entry tool (see Download section) or by following the detailed instructions below. Data needs to be in the correct format before processing through FETA. Please look at the completed example of the FFQ, which contains all the necessary codes to enter your FFQ data. Download the data entry template and use the worksheet "FETA template" which contains all the fields necessary for FETA. If you would prefer to enter data for the entire FFQ including fields not used by FETA, use the worksheet "Full FFQ Template" instead.

You might find it helpful to consult the worksheets 'sample data' and 'key' in the template spreadsheet. To ensure that processing goes smoothly, please keep to the format in the template and do not delete or rename compulsory column headings (all those in template coloured green).


The FFQ is separated into two parts - Part 1 (the Main Grid, pages 3-8) and Part 2 (back of the FFQ, pages 9-11).

The Main Grid

The frequencies are coded 1-9, from ‘Never or less than once/month’ (1) to ‘6+ per day’ (9), respectively as shown below:


Frequency

Never or less than once/month

1-3 per month

Once a week

2-4 per week

5-6 per week

Once a day

2-3 per day

4-5 per day

6+ per day

Data Entry Code

1

2

3

4

5

6

7

8

9

If two frequencies are given for 1 line, both may be entered, separated by a semi-colon, e.g. 2;3. Currently the program selects the first frequency value given.

If a line does not have a frequency, please enter -9 (which denotes missing data).


Back of the FFQ

Instructions will only be given on how to code the ticks and text that are processed by the program.

There are three look-up lists provided to translate the text for milk type (question 3), cereal types (question 5) and fat types (questions 6 & 7) into food codes, which have nutrient data attached to them.


Question 3 - MILK TYPE

See Figure 1 for the appropriate milk code to assign to each ticked box.

Where free text is given please enter the food code that best fits, using the milk look-up list. There is a non-specific milk code (a2002) available if a combination of milks is given or the text is uninformative. Currently the program only deals with one type of milk, so please do not enter more than one milk code.

The list of milk types does not contain recent additions to the market, such as rice/oat/coconut milks.


Question 4 - MILK QUANTITY

Please enter the appropriate amount code (MILK_FREQUENCY), as given in the example document. Please enter -9 if no box is ticked.

If an individual records that they consume milk in question 4, but do not specify a type in question 3, a default milk code (a2002) is assigned by the program.


Question 5 - CEREAL TYPE(S)

Please enter 1 for Yes or 2 for No (EAT_BREAKFAST); -9 if neither box ticked.

The text given must be translated into food code(s), using the cereal look-up list. You can list up to 4 food codes, separated by a semi-colon – please see the sample spreadsheet for examples. Do not worry about coding the brand information as this is not used by the program and will not alter the cereal code assigned.

Again, a non-specific cereal code (a1362) is available where the information provided can not be assigned a specific cereal code.

There is no cereal code provided for porridge in the look-up list. The reason for this is that porridge is dealt with in line 23 of the main grid, not line 24, and therefore should not be listed in question 5. If porridge is mentioned in this question, please ignore it.


Question 6 - FAT (frying, roasting, grilling)

This question asks about the type of fat used most often for frying, roasting, grilling etc. Please see Figure 2 for the appropriate fat code to assign to each ticked box, where appropriate. There is a fats look-up list to help you choose the most appropriate food code to match to the free text, where necessary (applicable to ‘vegetable oil’ and ‘none’ boxes only). Currently the program only deals with one type of frying fat, so please do not enter more than one frying fat code.

Question 7 - FAT (baking)

This question asks about the type of fat used most often for baking cakes etc. Please see Figure 3 for the appropriate fat code to assign to each ticked box, where appropriate. There is a fats look-up list to help you choose the most appropriate food code to match to the free text, where necessary (applicable to ‘margarine’ and ‘none’ boxes only). Currently the program only deals with one type of baking fat, so please do not enter more than one baking fat code.

A number of EPIC-Norfolk participants ticked the ‘none’ box for baking and frying fats. It was decided that it was more likely that fat was used but that the type was unknown. In the new program, if the ‘none’ box is ticked, the appropriate default fat is used (default frying code:17046, default baking code: 17018), in the absence of free text.


Question 10 - VISIBLE FAT

This question asks about the consumption of visible fat on meat (beef, pork, lamb, bacon and ham). It links back to lines 1, 3, 5, 7 and 9, respectively, in the main grid of the questionnaire, where each meat item consists of a portion of lean meat and a portion of separable fat. The program applies 100% of the separable fat amount if all fat is eaten, 50% for some fat eaten and 0% if 'ate as little fat as possible' or 'did not eat meat' were chosen.


Finished entering your FFQ data?

If you have entered your data using excel, once you have checked that it has all been entered as required, choose ‘save as’ and select the ‘CSV (comma delimited)’ type. Your dataset is now ready for FETA processing. Click on the following icon/link to install FETA. Please follow the instructions in the readme file that is created when the application is installed.

Once your input file is prepared, you can click here to access the download link(s).


PROCESSING OF FFQ DATA

Nutrient data and outputs available

The list of 46 nutrients and 14 food groups (default) calculated using FETA are available as: Default_Nutrients_FETA.xls. The nutrient data for the FFQ foods have come from McCance and Widdowson’s The Composition of Foods (5th edition) and its associated supplements (see reference section).

There are four types of output available for nutrient intake data. The default output is "Output 1".

Output 1 contains average daily nutrient intake for an individual from all FFQ foods consumed. It is the most helpful format for analysis and suitable for import into a spreadsheet or statistical software.

Table 1. Nutrient data for an individual from all foods consumed.


ID

Calcium

Cholesterol

Chloride

Copper

mg

mg

mg

mg

1234567A

1327

412.5

8156

1.757

Output 2 contains the same data as output 1 (average daily nutrient intake for an individual). The only difference is that the data in this output is in long format, as opposed to wide (spreadsheet) format in output 1.

Table 2. Nutrient data for an individual from all foods consumed.


ID

NUTRIENT_CODE

NUTRIENT_QUANTITY

1234567A

8

1327

1234567A

12

412.5

1234567A

13

8157

1234567A

15

1.757

1234567A

21

113.2

1234567A

22

16.13

1234567A

55

438.3

1234567A

61

234.8

1234567A

63

5067

Output 3 contains average daily nutrient intake at the meal_id or row level for an individual (Table 3). It also contains the amount of food consumed in each meal_id. A description of the MEAL_ID is to be found in the 'MEAL-ID look-up' worksheet in the data entry template. In this output, the nutrient quantities and meal portions in output 4 are summed for the two food codes, to provide average daily nutrient data at the meal_id or row level.

Table 3. Nutrient data for an individual at the meal_id/row level.


ID

MEAL_ID

MEAL_PORTION

NUTRIENT_CODE

NUTRIENT_QUANTITY

1234567A

5

49.45

8

5.450

1234567A

5

49.45

12

51.6

1234567A

5

49.45

13

37.73

1234567A

5

49.45

15

0.0488

1234567A

5

49.45

21

3.522

1234567A

5

49.45

22

0.3526

1234567A

5

49.45

55

5.397

1234567A

5

49.45

61

3.526

1234567A

5

49.45

63

164.5

The most detailed output, shown in Table 4, contains average daily nutrient intake at the food code level for an individual. It also contains the amount of food consumed. The meal_id is the row number in the main grid; row 5 is chicken or other poultry, e.g. turkey.

Table 4. Nutrient data for an individual at the food level


ID

MEAL_ID

FOOD_CODE

FOOD_PORTION

NUTRIENT_CODE

NUTRIENT_QUANTITY

1234567A

5

18331

43

8

4.73

1234567A

5

18331

43

12

45.15

1234567A

5

18331

43

13

32.25

1234567A

5

18331

43

15

0.043

1234567A

5

18331

43

21

3.225

1234567A

5

18331

43

22

0.301

1234567A

5

18331

43

55

4.3

1234567A

5

18331

43

61

3.01

1234567A

5

18331

43

63

141.9

1234567A

5

18361

6.45

8

0.709

1234567A

5

18361

6.45

12

6.45

1234567A

5

18361

6.45

13

5.483

1234567A

5

18361

6.45

15

0.0058

1234567A

5

18361

6.45

21

0.297

1234567A

5

18361

6.45

22

0.0516

1234567A

5

18361

6.45

55

1.096

1234567A

5

18361

6.45

61

0.516

1234567A

5

18361

6.45

63

22.57

Detailed data at the food level has a number of uses. In addition to the average daily nutrient intake at the food level, the average daily consumption of each food item is also provided. The average daily amounts of food consumed could be combined as required to provide food group data, either using outputs 3 or 4.


You can email any queries/feedback you might have to angelam@srl.cam.ac.uk. We will try and answer these as soon as we can.



Home | About | Third Health Check | International | People & Galleries | Participant Advisory PanelJobs | Contact Us | FAQ |

Findings & ResultsPublications | In the News | Collaborators | Nutritional Methods | Public Events | Sponsors | Intranet | Local |

last updated: 20 March 2013