• Indication
• Pros and cons
• Practical considerations
• Resources required
• References
• Examples and links

• Dietary assessment
• Methods
• Food frequency questionnaire
• Resources required

Resources required

1. Trained interviewers (if the FFQ is interviewer administered)
Interviewer-administered FFQs will help ensure that questionnaires are fully completed and should help reduce any ambiguity respondents have with specific questions. Face-to-face interviews or telephone interviews can be carried out. Interview by telephone can be less costly than in-person interviews and have the advantage of being able to reach large number of people in different geographic regions. It is also very time consuming and tedious to answer long lists of questions in the FFQ format if it is being interviewer administered, requiring full concentration to the end. Appropriate ways of asking the questions and recording responses need to be considered, not just reading out the questionnaire. If young children or adults with incapacity are being studied, the parent/guardian or carer should be present. Additional information may also be gained from caterers such as schools or care home cooking staff.   

2. Portion estimation aids
Posting picture booklets or other portion size estimation aids to respondents can help simplify the reporting of portion sizes. Colour photographs of key portions can also be included in the FFQ. The estimation of portion size in the recall may be improved with the use of aids such as food photograph atlas, household measures or food models.

3. An appropriate nutrient database
To convert frequency estimates of food intake to nutrient values use of an appropriate nutrient database is required. For each food or drink listed in the FFQ, the nutrient composition is selected, though this can be an average of the composition of several different foods or varieties of the same food (e.g. 'pizza' nutrient composition may be defined as the average for some or all of the different entires for pizza in the parent food compostion tables). More complex calculations are sometimes used e.g. use of the specific type of cooking oil in calculating the fat content for all home-ccoked foods can improve the estimation of fatty acid intake.
FFQs that are computer-readable may help eliminate data-entry errors and reduce the time required for analysis though on-screen checking of entries is recommended as hand-written responses may not always be detected accurately.

4. Standardised operating procedures - for data checking, cleaning and analysis.

Back to top