Dietary Data Collection: Food Frequency Questionnaire

Subject: Healthcare Research
Pages: 7
Words: 2032
Reading time:
9 min
Study level: PhD


In the years between the 1950s and 1960s, nutritionists started developing a questionnaires as a technique for assessing habitual food intake based on recorded food consumed over a set time period because of the increasing health problems associated with nutritional risk factors in the elderly, adolescents, pregnant and adult population. FFQ research is carried out by comparing the adequacy of the diet of the respondent is currently undertaking such as the dietary reference intakes (DRI’s) by drawing dietary assessment, nutritional deficiencies, or excesses. The research done relates to food consumption habits and diet patterns relating to the respondent that has actuated food frequency questionnaires (FFQ’s) to collect long-term dietary data. The method we are going to employ here in the dietary assessment will enable us to recall surrogate respondents’ food intakes over the past 6 months-one year.


Tools commonly used for assessment to evaluate food intake include Food Frequency Questionnaires (FFQ) and contain about 100 different foods. The research carries out aims to recall intake for foods consumed for a refined reference period of over one year. The participant in this experiment requires to recollect food intake during a certain period of time combined with regular eating habits. FFQ method of diet assessment is frequently used in epidemiological research on diet and diseases to help classify individuals according to their usual intake. Older persons have been frequently used in this dietary assessment due to their varying degrees of cognitive dysfunctions. (Tamara, 2003, p.13).

Methods used

Food frequency approach, the participants are asked to report their regular intake of each food for a specific period of time. FFQ assessment incorporates portion size, methods of cooking, what makes a combination in a meal in order to estimate relative/absolute nutrients consumed on a daily basis of an individual’s diet. To determine all nutrients, FFQ sums up all food products frequency of every food by the number of nutrients in that particular serving to come up with an estimate of daily intake of nutrients, what constitutes the diet, and classify food groups of a particular person (Coulston & Boushey, 2008, p.7).

FFQ is performed by asking a respondent to give a detailed report of foods consumed frequently. The report will include giving an account of food eaten alone, mixed in other foods, or meals consumed separately. For example in beans, it could be served alone or in a mixture form. In carrying out the assessment, separate questions could be asked about refried beans, beans soups, and beans mixed in burritos. In compiling a report basing on these questions, the researcher needs to double count the food where beans were mixed in burritos as a Mexican mixture and beans alone. In other cases, FFQs contain a similar group of foods in one question, beef, pork, lamb, will be classified under one name such as meat. Classifying food under one group can be challenging because it’s hard to apply one question to a group of foods that vary in their nutritional content.

In order to determine each quantitative FFQ, base data must be incorporated to it to effectively estimate nutrients intake for each portion size of food reported. Preparation of foods such as macaroni and cheese, different people use different recipes each entailing different nutrients composition encompassed to the meal yet the FFQ database requires to have a profile containing a single nutrient composition. In order to come up with a clear database in such a case, FFQ uses quantitative dietary intake data from a group of individual like children, adults or old people to define the average nutrients quantity of the chosen group. All reports pertaining results of different food groups such as macaroni and cheese of different individuals collected in a population survey, results should attained by calculating mean or median to reach the nutrients composition estimates. In order to obtain more accurate results, dietary analysis software may be used such as the FFQ to compute nutrients intake for the respondents. (Coulston &Boushey, 2008, p.7).

Food frequency questionnaire investigators are putting efforts to improve the validity of the frequently questionnaire design issues such as portion size, seasons, length, closed and open ended questions and time framed for the questionnaires. The current problem associated with FFQ instruments formulated to assess total diet usually list more than 100 items linked with many extra portion size questions which requires a lot of time to of about 30 to 60 minutes to complete. The long duration of time required to complete a particular question affect the validity of the response. It becomes difficult to define a closed-ended questionnaire for a list of foods used for food frequency instruments. With the birth of new technology such as scanning instruments has contributed to the loss in specificity of the information reported therefore necessitating closed-ended responses in questionnaires (Tylavsky & Sharp, 1995, p.349).

Amount of foods consumed are important component in estimating dietary intakes, however, FFQ instrument does not clearly indicate whether portions size questions should be included or not. Frequency method of data collection has been commonly used in the typical serving size of determining food intake. In order to improve the performance of FFQs, the researcher should incorporate portion size and frequency into one question either than the long detailed questions. Time frame is another way to improve FFQ performance is by reducing the period in which research is conducted. The initial period of one year could reduce to one month or past one week to cut administrative cost and provide more reliable results (Coulston &Boushey, 2008, p.8).

Advantages of using FFQ as a dietary assessment tool

Unlike other dietary assessment tools, FFQ method is less costly in administering and processing reports and respondent. Its objective is to estimate the participant’s regular food intake over a period of time. Eating behavior is not altered therefore FFQ does not report recent changes in diet resulting from either illness or short travels. Obtaining information regarding a person’s diets is recalled from a previous time period and therefore relative low cost of carrying out research. Nearly most of the food frequency instruments take less time of about 30-60 minutes to complete and in some cases the responses are optical scanned paper version therefore lowering the burden and cost of data collection than the methods used for recalls (Nutrition Quest, 2007) (Nutrition Assessment, 2007).


It lacks detailed description of foods eaten due to the various food brands, preparation practices and many foods entail different nutritional contents. It does not clearly describe cooking methods involved in preparing the food and brand names of drinks and foods taken. FFQ method of assessment is complex in a way that requires higher mathematical knowledge in order to be able to clearly calculate the food that is less frequently consumed. The participant is also required to have calculations knowledge to be able to determine the portions he or she consumed and be able to read the list. Obtaining accurate report form the assessment about foods eaten both as single and in mixtures becomes problematic (Coulston &Boushey, 2008, p.7).

The other major limitation is the measurement error due to numerous calculations it entails. When carrying out FFQ, many dietary intakes are not measured therefore the quantifications do not reflect the records. Incomplete listings of foods codes and errors attributed from the frequency and estimations of portion size contribute to inaccurate results. FFQ estimation tasks are often complex and difficult. In normal circumstances, a person does not take the same nutrients all the time, it changes considerably therefore resulting to inaccurate estimates on a person’s diet. Longer food frequently samples list collected over a long period of time amounts to overestimates of the average intake of a particular group results. On the other hand, shorter lists may underestimate the nutrients intake results due to varying degrees of each foods nutrient such as vegetables and fruits can not be adequately estimated in terms of their nutritional content (Smith, 1993, p. 11) (Heimendinger, Krebs-Smith et al, 1994).

Researchers claim that it’s inappropriate to use FFQ data for estimating quantitative parameters because of the congenital error relating to the FFQ approach such as the use of mean and variance to base population dietary intake. The data obtained from FFQ nutrients intake should only be considered as only estimates and not actual data because it uses different population estimates with varying nutritional intakes. FFQ ranks topics to be researched on according to nutrients and food intake either than estimating food absolute levels of nutrients intake which are used to assess the connection between dietary intake and disease risk thereby providing preventive measures or cure to some diseases associated with it (Block & Subar,1992, p. 969-977) (Coulston &Boushey, 2008, p.8).

The actual validity study of food frequency-based on estimates of long-term diet assessment would require non disrupted study of the respondent’s total diet over a long period of time and from the statistics, no such studies has ever been done. A more practical approach in estimating food frequency would be multiple food recalls and reports can be drawn over a period of time as an indicator of normal diet. This method clearly examines the relationship between foods and nutrients that range from 0.4-0.7 that draws its reports from, however this method of recall can not be considered as accurate reference as they also prone to error in the FFQ and it may not accurately represent the time period in which the results are drawn from (Schaefer, Dallal et al, 2000).

Validation of the FFQ is a very important method of assessment as it determines the degree to which the questionnaire carried out is formulated in order to adequately measure food or nutrients. Without the validation, incorrect information recorded may lead to false relationship represented between dietary factors and diseases or false alarms to disease-related markers (Cade et al, 2004, p.6).

Before food frequency questionnaire is used as a data collection method, frequency and homogeneity consumption in the source population should be considered. Other factors such as sex, age and education should be kept in mind. Relating to the previous research based on reproducibility of a semi-quantitative index drawn, the recall of adolescent diet has been seen to have a low validity (Cat. Inist, 1988). FFQ are most considered the most appropriate method of assessing diet over other methods. Researchers choose to use it because it provides better assessment results due to the longer period it takes rather than one day method. This is important in studying pregnant women because their diet do not remain constant and therefore assessment of past month results may be more accurate than during the past one year (Blum, Rockett et all, 2005) (Kristal, Feng et al, 1997).

Cancer prevention trials are using food frequency questionnaire to help in eligibility screening for patients whose daily calcium intake is less that 500 mg per day for safety reasons and those whom dietary fiber is greater than 30g per day. A study was done to assess the final eligibility determinations. The subject picked out was Wheat Bran Fiber Trial to evaluate the sensitivity, specificity the negative and positive predictive value on the food frequency questionnaire. 183 members participated in the assessment and a 4-day food records were collected. Calculations were done on negative prediction value basing on the results collected from the four day food records grouping men and women separately targeting calcium and dietary fiber values using interim analysis. When the health profession note the decline or increase in calcium levels of both men and women, they are able to advice preventive measures or medications well in advance (Ritenbaugh, Aickin et al, 1997).

In conclusion, if short term assessment of 6 months to one week procedure is used rather than the one year period, future research will improve the quality of data collected and the results relied upon by health professions in diagnosing diseases. Every questionnaire should be judged according to the provided information it was intended for as there is no standard FFQ measurement tool to base all the results from. Censors method of data collection has been incorporated in FFQ technique to assist in planning and analyzing information collected in order to optimize its performance.


Block, G., & Subar, A. F. (1992). Estimates of nutrient intake from a food frequency questionnaire: The 1987 National Health Interview Survey. J. Am Diet Association, 969-977.

Blum, R. E., Heather, J, B., Rockett, H. R., Leppert, J, Garner, J. D., Suitor, C. W., & Graham, A. C. (2005). Use of a food frequency questionnaire in American Indian and Caucasian pregnant women: a validation study. 2008. Web.

Cat.Inist (1988). Factors affecting reproducibility of diet reports using food frequency Questionnaires. Web.

Cade, J.E., Burley, V. j., Warm, D.L., Thompson, R.L., & Margetts, B. M. (2004). Food-frequency questionnaires: a review of their design, validation and utilization. Nutritional Research Reviews, 17, 5-22.

Coulston, M.N., & Boushey, C. J. (2008). Nutrition in the Prevention and Treatment of Disease. Academic press.

Kerbs-Smith, S.M., Heimendinger, J., Subar, A, F., Patterson, B. H., & Pivonka, E. (1994). Estimating fruit and vegetable intake using food frequency questionnaire: A comparison of instrument. Am, J. Clinical Nutrition, 59, 283.

Kristal, A. R., Feng, Z, Coates, R. J., & George, V. (1997). Association of race/ethnicity of a food frequency questionnaire: the Women’s Health Trial Feasibility Study in Minority Populations. Am J Epidemiol, 146, 856-859.

Ritenbaugh, C., Aickin, M., Taren, D., Teufel, N., Grave, E., & Albert D.S. (1997). Use of food frequency questionnaire to screen for dietary eligibility in a randomized cancer prevention phase III trail. Cancer Epidemiol Biomarkers, 6, 347-354.

Nutrition Quest: Questionnaire &screeners. (2007). Nutrition Quest. Web.

Mattys, C., Pynaert, I., De Keyzer, W., & De Henauw, S. (2007). Validity and reproducibility of adolescent web-based food frequency questionnaire. J. Am. Diet Association, 107, 605-610.

Nutrition assessment. (2007). Fred hutchinson Cancer Research Center. Web.

Schaefer, E. J., Augustin, J., L., Schaefer, M. M., Rasmussen, H., Ordovas, J.M., Dallal, G. E., & Dwyer, J. T. (2000). Lack of efficiency of a food-frequency questionnaire in assessing dietary macronutrient intakes in subjects consuming diets of known composition. Am. J. Clinical Nutritional, 71, 746-751).

Smith, A.F. (1993). Cognitive psychological issues of relevance to the validity of dietary reports. Eur. J. Clinical Nutrition, 47, 6-18.

Tamara, A. H. (2003). Validation of Multi-call 24-Hour Recall in Older Adults with cognitively impairments. may, 2008, 1-44.

Tylavsky, F. A., &Sharp, G.B. (1995). Misclassification of nutrient and energy intake from use of closed-ended questions in epidemiologic research. Am. J. Epidemiol, 142, 342-352.