![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
German Institute of Human Nutrition Potsdam-Rehbruecke, 14558 Nuthetal, Germany
* To whom correspondence should be addressed. E-mail: ute.noethlings{at}dife.de.
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
Introduction |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
The prospective design of studies using active follow-up procedures provides a unique opportunity to obtain several measurements of diet over time. The use of repeated dietary measurements has been shown to provide stronger associations and narrower CI than single measurements (8). However, comprehensive dietary assessment is a methodological challenge, because total and partial nonresponse to FFQ have to be largely precluded to minimize loss to follow-up. A dietary assessment instrument minimizing respondent burden therefore seems advantageous.
For the repeated dietary assessment in the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam Study, a simple FFQ with low respondent burden was developed. To improve FFQ methodology, 2 different concepts to implement portion sizes were applied and compared regarding measurement of absolute intake and ranking of study participants.
|
Materials and Methods |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
We adopted a semiquantitative style, assessing frequencies of food intake but not individual portion sizes. Previously, our group showed that questions about individual portion sizes were of minor importance to assess variation in intake (17). Accordingly, participants were asked to report the frequency of consumption of 1 graphically displayed portion for most food items. Furthermore, for the simple FFQ, we adopted a nongrid format with questions and answers that appeared as a unit for every food item. Research has suggested this format is cognitively easier for respondents (18,19). In a pilot study including 512 EPIC-Potsdam study participants, 94% of the simple FFQ were returned with 10 or fewer missing responses. During the subsequent follow-up of the total cohort in which the simple FFQ was applied, a response proportion of 93% was achieved.
Study design. At baseline between 1994 and 1998, 27,548 participants were recruited to the EPIC-Potsdam study (20–22). Active follow-up using questionnaires is conducted approximately every 2 y (23). In 2004, 5673 study participants were scheduled to receive the simple FFQ. The questionnaire also inquired about current weight, height, and smoking status. For validation purposes, we used a random sample of 230 men and 230 women to provide 2 unannounced 24HDR, randomized over the year preceding administration of the FFQ (i.e. reflecting consumption habits during the time period also covered by the FFQ). We decided to use 2 24HDR rather than a larger number to minimize the interference with the routine follow-up and thereby decrease the chance of withdrawals from the study. Eighty-eight percent (n = 406) of the participants agreed to answer a computer-assisted 24HDR (EPIC-Soft) (24) via telephone. A total of 393 participants (197 men and 196 women) completed both interviews and the FFQ at the end of the year. Missing answers in FFQ were identified and participants were contacted via telephone to complete all information.
Determination of portion sizes. Daily food intake in grams was calculated for each study participant based on the frequency reported in the FFQ and a portion size, which we assigned to food items. We compared 2 different approaches in assigning the portion size: a predefined portion size based on prior knowledge and a fitted portion size based on the 24HDR, which covered the same time period as the FFQ in this validation study. For predefined portion sizes, amounts were used that have already been used for the implementation of the baseline FFQ, or, for mixed dishes, median intakes in baseline 24HDR. To define fitted portion sizes, the average daily intake of each food item consumed in the recent 2 24HDR was summed separately for all men or women and divided by the summed frequencies of consumption per day reported in the simple FFQ. Therefore, by definition, the mean of the product of fitted portion size and frequency of intake approximated the mean intake measured in the reference 24HDR for each food item for men and women. Accordingly, based on our previous findings that sex-specific portion sizes provided little information on variation in intake, the predefined portion sizes were identical for men and women for each food item, whereas the fitted portion sizes differed between sexes.
Statistical analysis. We calculated the mean intakes and variances for the nutrients and food groups. We decomposed the variance assessed in the 2 reference 24HDR into inter- and intra-individual variance and report the SD of the mean derived from inter-individual variance along with the mean. To evaluate the ranking of study participants, Pearson correlation coefficients (r) were calculated, deattenuated (i.e. corrected for intra-individual variance) (25), and energy adjusted applying the residual method (26). Furthermore, food items have been grouped to key food groups for each nutrient by sex. For each nutrient, food items that had explained 80% or more of variance in nutrient intake in the baseline FFQ, as determined during the development of the new FFQ, were grouped. Subsequently, nutrient intakes were calculated based on their respective key food items only and deattenuated r with the 24HDR measurement were calculated.
All analyses were performed using SAS (version 9.1, SAS Institute).
|
Results |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
|
|
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
In general, FFQ alone were considered to obtain suboptimal quantitative measurements of intake. The use of standardization studies has been suggested to provide an estimate of the magnitude of dietary measurement error by providing a quantitative assessment of food consumption (27–29). As shown, our FFQ alone, i.e. without the additional information of the reference 24HDR used to fit portion sizes, provided suboptimal measures of absolute intake too. Absolute intakes of food groups and nutrients were much more reasonable compared with our reference measurement after implementation of fitted portion sizes compared with predefined portion sizes. Although our instrument was designed to reflect the ranking of study participants, we also tried to obtain reasonable absolute intakes, because we think this will ease the communication of research findings. The ranking regarding single nutrients was affected because fitted portion sizes differed for men and women, but effects were marginal only and correlation coefficients with fitted portion sizes did not generally improve.
The fitted portion sizes could be interpreted as weighting factors for food items, because food items were used to derive nutrient and broader food group intake. The use of the fitted portion sizes resulted in proportions of food items in broader food groups or nutrients that reflect proportions assessed in the reference data.
Our selection of the predefined portion sizes was based on available information, i.e. dietary assessments in EPIC-Potsdam, and could therefore be argued as being arbitrary. If different amounts would have been specified a priori, different absolute intakes of food groups and nutrients would have been calculated. During the development of the questionnaire, however, predefining portion sizes following earlier definitions was the best possible procedure at that point, although these portion size estimates were more than 10 y old. There is evidence in the literature that usual portion sizes increased over time (30), which could have affected our comparison. However, our study also suggests the use of recent reference data to specify portion sizes in the FFQ.
Correlation coefficients for foods and nutrients in other validation studies often ranged between 0.4 and 0.7 (18). The correlation coefficients for our instrument mostly were within this range but generally moderate in magnitude. We also investigated if correlation coefficients for nutrients solely based on key food items were higher than correlation coefficients based on all food items. Crude correlations were higher for some nutrients, but after deattenuation and energy adjustment, this approach did not provide a better ranking of study participants. Overall, our simple FFQ is of reasonable validity for use in epidemiological studies. However, despite careful design, we were not able to derive an FFQ with much better measures of validity than FFQ usually obtain. We therefore are in favor of the concept of a ceiling of validity for FFQ (31) and suggest that future research explores other dietary assessment methods or combinations of methods for use in large epidemiological studies.
A limitation of our study is the reference dietary measurement. We used repeated 24HDR measurements for the validation study based on the assumption that the 24HDR provide a reasonable quantitative assessment of dietary intake. However, besides the fact that errors in both instruments are considered correlated due to their retrospective nature, it has been shown that 24HDR largely underestimate dietary intake when compared with the gold-standard biomarkers for energy (doubly labeled water) and protein (urinary nitrogen) (3,32). Furthermore, we obtained only 2 24HDR for each study participant, which on the one hand resulted in high within-person variation in the reference measurement and on the other hand might have impaired the assessment of seasonal variation. However, we accounted for within-person variation in our statistical analysis and the randomized sampling procedure for the 24HDR resulted in an almost equal distribution across seasons.
A strength of our study was the assessment of 2 24HDR that allowed us to obtain an estimate of intra-personal variation. Eliminating the intra-personal variance component yielded an estimate of inter-personal variance that reflected variation of individuals' usual intake. Thus, we were able to compare the observed variance of FFQ data with variation of long-term dietary intake expected from our reference measurements.
In conclusion, the use of recent reference data from a validation study to fit portion sizes was useful to achieve dietary data with an FFQ interpretable on an absolute scale. The ranking of study participants for nutrient or food group intakes could not be improved by fitting portion sizes. We consider our simple FFQ with fitted portion sizes a cost-effective instrument of reasonable validity with the benefit of ease of use in large epidemiological studies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Author disclosures: U. Nöthlings, K. Hoffmann, M. M. Bergmann, and H. Boeing, no conflicts of interest.
Manuscript received 21 June 2007. Initial review completed 30 July 2007. Revision accepted 2 October 2007.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
1. Freedman LS, Potischman N, Kipnis V, Midthune D, Schatzkin A, Thompson FE, Troiano RP, Prentice R, Patterson R, et al. A comparison of two dietary instruments for evaluating the fat-breast cancer relationship. Int J Epidemiol. 2006;35:1011–21.
2. Bingham SA, Luben R, Welch A, Wareham N, Khaw KT, Day N. Are imprecise methods obscuring a relation between fat and breast cancer? Lancet. 2003;362:212–4.[Medline]
3. Subar AF, Kipnis V, Troiano RP, Midthune D, Schoeller DA, Bingham S, Sharbaugh CO, Trabulsi J, Runswick S, et al. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol. 2003;158:1–13.
4. Willett WC, Hu FB. The food frequency questionnaire. Cancer Epidemiol Biomarkers Prev. 2007;16:182–3.
5. Willett WC, Hu FB. Not the time to abandon the food frequency questionnaire: point. Cancer Epidemiol Biomarkers Prev. 2006;15:1757–8.
6. Kristal AR, Peters U, Potter JD. Is it time to abandon the food frequency questionnaire? Cancer Epidemiol Biomarkers Prev. 2005;14:2826–8.
7. Kristal AR, Potter JD. Not the time to abandon the food frequency questionnaire: counterpoint. Cancer Epidemiol Biomarkers Prev. 2006;15:1759–60.
8. Hu FB, Stampfer MJ, Rimm E, Ascherio A, Rosner BA, Spiegelman D, Willett WC. Dietary fat and coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Am J Epidemiol. 1999;149:531–40.
9. Boeing H, Wahrendorf J, Becker N. EPIC-Germany: a source for studies into diet and risk of chronic diseases. European Investigation into Cancer and Nutrition. Ann Nutr Metab. 1999;43:195–204.[Medline]
10. Bohlscheid-Thomas S, Hoting I, Boeing H, Wahrendorf J. Reproducibility and relative validity of energy and macronutrient intake of a food frequency questionnaire developed for the German part of the EPIC project. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol. 1997;26 Suppl 1:S71–81.
11. Bohlscheid-Thomas S, Hoting I, Boeing H, Wahrendorf J. Reproducibility and relative validity of food group intake in a food frequency questionnaire developed for the German part of the EPIC project. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol. 1997;26 Suppl 1:S59–70.
12. Kroke A, Klipstein-Grobusch K, Voss S, Moseneder J, Thielecke F, Noack R, Boeing H. Validation of a self-administered food-frequency questionnaire administered in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study: comparison of energy, protein, and macronutrient intakes estimated with the doubly labeled water, urinary nitrogen, and repeated 24-h dietary recall methods. Am J Clin Nutr. 1999;70:439–47.
13. Slimani N, Kaaks R, Ferrari P, Casagrande C, Clavel-Chapelon F, Lotze G, Kroke A, Trichopoulos D, Trichopoulou A, et al. European Prospective Investigation into Cancer and Nutrition (EPIC) calibration study: rationale, design and population characteristics. Public Health Nutr. 2002;5:1125–45.[Medline]
14. Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, Hennekens CH, Speizer FE. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122:51–65.
15. Byers T, Marshall J, Fiedler R, Zielezny M, Graham S. Assessing nutrient intake with an abbreviated dietary interview. Am J Epidemiol. 1985;122:41–50.
16. Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol. 1986;124:453–69.
17. Noethlings U, Hoffmann K, Bergmann MM, Boeing H. Portion size adds limited information on variance in food intake of participants in the EPIC-Potsdam study. J Nutr. 2003;133:510–5.
18. Subar AF, Thompson FE, Kipnis V, Midthune D, Hurwitz P, McNutt S, McIntosh A, Rosenfeld S. Comparative validation of the Block, Willett, and National Cancer Institute food frequency questionnaires: the Eating at America's Table Study. Am J Epidemiol. 2001;154:1089–99.
19. Subar AF, Ziegler RG, Thompson FE, Johnson CC, Weissfeld JL, Reding D, Kavounis KH, Hayes RB. Is shorter always better? Relative importance of questionnaire length and cognitive ease on response rates and data quality for two dietary questionnaires. Am J Epidemiol. 2001;153:404–9.
20. Riboli E, Hunt KJ, Slimani N, Ferrari P, Norat T, Fahey M, Charrondiere UR, Hemon B, Casagrande C, et al. European Prospective Investigation into Cancer and Nutrition (EPIC): study populations and data collection. Public Health Nutr. 2002;5:1113–24.[Medline]
21. Riboli E, Kaaks R. The EPIC Project: rationale and study design. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol. 1997;26 Suppl 1:S6–14.
22. Boeing H, Korfmann A, Bergmann MM. Recruitment procedures of EPIC-Germany. European Investigation into Cancer and Nutrition. Ann Nutr Metab. 1999;43:205–15.[Medline]
23. Bergmann MM, Bussas U, Boeing H. Follow-up procedures in EPIC-Germany: data quality aspects. European Prospective Investigation into Cancer and Nutrition. Ann Nutr Metab. 1999;43:225–34.[Medline]
24. Slimani N, Ferrari P, Ocke M, Welch A, Boeing H, Liere M, Pala V, Amiano P, Lagiou A, et al. Standardization of the 24-hour diet recall calibration method used in the european prospective investigation into cancer and nutrition (EPIC): general concepts and preliminary results. Eur J Clin Nutr. 2000;54:900–17.[Medline]
25. Beaton GH, Milner J, Corey P, McGuire V, Cousins M, Stewart E, de Ramos M, Hewitt D, Grambsch PV, et al. Sources of variance in 24-hour dietary recall data: implications for nutrition study design and interpretation. Am J Clin Nutr. 1979;32:2546–59.
26. Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986;124:17–27.
27. Fraser GE. A search for truth in dietary epidemiology. Am J Clin Nutr. 2003;78:S521–5.
28. Kushi LH. Gaps in epidemiologic research methods: design considerations for studies that use food-frequency questionnaires. Am J Clin Nutr. 1994;59:S180–4.
29. Hoffmann K, Kroke A, Klipstein-Grobusch K, Boeing H. Standardization of dietary intake measurements by nonlinear calibration using short-term reference data. Am J Epidemiol. 2002;156:862–70.
30. Nielsen SJ, Popkin BM. Patterns and trends in food portion sizes, 1977–1998. JAMA. 2003;289:450–3.
31. Willett W. Invited commentary: a further look at dietary questionnaire validation. Am J Epidemiol. 2001;154:1100–2.
32. Kipnis V, Subar AF, Midthune D, Freedman LS, Ballard-Barbash R, Troiano RP, Bingham S, Schoeller DA, Schatzkin A, et al. Structure of dietary measurement error: results of the OPEN biomarker study. Am J Epidemiol. 2003;158:14–21.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
