Food frequency questionnaires (FFQ) are self-administered instruments that obtain data on frequency and portion size on anywhere from 60 to 150 foods or food groups. Respondents indicate on the form whether they eat a certain food and how often (i.e., once a month or several times a day). These instruments usually include various types of questions: (1) adjustment questions on food preparation or frequency of restaurant eating, (2) the food list, and (3) summary questions. The adjustment questions permit more refined analyses of fat intake by asking about food preparation practices (e.g., removing skin from chicken), types of added fats (e.g., use of butter vs. margarine on vegetables), or type of milk usually used (e.g., whole, skim, soy). Because details about food preparation are not obtained with FFQ in the same manner as food records and recalls, these adjustment questions can be useful approaches to fine-tune the food selections.
The main section of an FFQ consists of a food or food group list, with questions on usual frequency of intake and portion size. To allow for machine scanning of these forms, frequency responses are typically categorized from "never or less than once per month" to "2+ per day" for foods and "6+ per day" for beverages. Portion sizes are often assessed by asking respondents to mark "small," "medium," or "large" in comparison to a given medium portion size. However, some questionnaires only ask about the frequency of intake of a "usual" portion size (e.g., 3 ounces meat). Often a cartoon-type picture or very small photograph is provided as a "reference point" for the medium portion size. However, since the pictures must be small to fit onto the questionnaire, it is unclear how useful they may be in the long run.
FFQ food lists or line items are created to capture data on (1) major sources of energy and nutrients in the population of interest, (2) between-person variability in food intake, and (3) specific scientific hypotheses (Willett, 1998). Since the food list possibly cannot include the universe of all food and food combination possibilities, decisions must be made about which foods to include and which ones to omit. For example, one approach is to use national consumption data from 24-h dietary recalls to determine the most commonly eaten foods and major nutrient sources in the diet (Block et al, 1994, 1986; Willett, 1998). Details about nutrients are limited though for foods consumed in specific population groups (e.g., certain ethnic foods) and there are limited data on bioactive constituents of foods that are not considered nutrients but nonetheless have biological actions (e.g., isoth-iocyanates, polyphenols). Finally, to save space and reduce respondent burden, similar foods are often grouped into a single-line item (e.g., apples, peaches, plums). When grouping foods, important considerations include whether they are nutritionally similar enough to be grouped and whether the group will make cognitive sense to the respondent. For example, a food group composed of three types of fruit may have similar amounts of vitamin C, but be very different for other bioactive compounds such as flavonoids or carotenoids. These issues must be considered when selecting a dietary assessment instrument.
Was this article helpful?