![[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}
|
|
|
|
|
||
,**,2
,**,**,**
* Harvard School of Public Health, Departments of Nutrition and Epidemiology, Boston, MA;
S. Daniel Abraham International Center for Health and Nutrition, Ben-Gurion University of the Negev, Beer-Sheva, Israel;
** Epidemiology and Health Services Evaluation Department, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel;
Channing Laboratory, Harvard Medical School, Boston, MA; and
The Nuclear Research Center, Department of Medicine, Dimona, Israel
2To whom correspondence should be addressed. E-mail: ishai{at}hsph.harvard.edu and irish{at}bgumail.bgu.ac.il.
 |
ABSTRACT |
|---|
 TOP ABSTRACT SUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
-tocopherol (0.44 and 0.34, respectively). The DEARR study showed that the newly developed FFQ is a valid and reproducible instrument for assessing dietary intake. The VCs obtained can be used for future adjustment of diet-disease RR estimates in this population.
KEY WORDS: • biomarkers • diet • dietary questionnaires • reproducibility and validity • true intake
The current debate concerning the performance of FFQs (1–3) demonstrates the need for calibration studies to evaluate the factors required for deattenuating estimated diet-disease relative risks (RRs),3 especially for newly developed dietary assessment tools. The common calibration method for RR in dietary studies uses a multivariate regression or Cox proportional hazard regression, which is a linear approximation method for logistic regression (4). Reference methods such as multiple 24-h dietary recalls (24HRs) (5) or multiple-day food records (4) are used to adjust for attenuation. However, because the corrected RR is biased when the sources of error for the reference method and the tested tool are similar, (6,7) the addition of biomarkers can determine the lower limit for the validity coefficient (VC).
Kaaks and Ocke (8,9) suggested using the "method of triads," by which correlation (VC) of the dietary assessment method and "true" long-term intake could be estimated from 3 pairwise correlations between the FFQ, the reference method (24HRs or diaries), and the biomarker. This technique is an application of a factor analysis model and corrects for bias due to correlated errors in the repeated measurements from the reference method.
Although the performance of FFQs in estimating intake of individual nutrients has been evaluated widely in European countries and in the United States, (1,2,10,11), little is known about its performance in other populations, such as Mediterranean countries in the Middle East, compared with that of 24HRs. In the Dietary Evaluation and Attenuation of Relative Risks (DEARR) study, we evaluated the VCs of protein, folic acid, -tocopherol, and ß-carotene by integrating repeated measurements of a newly developed FFQ, 24HRs, blood drawings, and 24-h urine collections. Using the "method of triads," we also evaluated the performance of the urine and blood biomarkers in association with the "true intake" compared with the dietary assessment tools. The DEARR study joins the few published studies that have applied the "method of triads" (9,12,13).
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
35 y old, were recruited from a large research center. No restrictions, except that of pregnancy, were enforced. Over a period of 13 mo (Table 1), the participants were asked to complete the self-administered FFQ 3 times. Participants were interviewed in their homes on 6 occasions to obtain 24HRs. Three 24-h urine collections and 2 blood samples were obtained, and 3 clinical measurements were carried out in the participants’ workplace. To encourage better compliance, we offered the participants a detailed dietary assessment, a blood profile, and specific individually tailored dietary guidelines at the end of the project. We also rewarded the volunteers with a gift after each step of the study. The University board (Ben-Gurion University of the Negev, Beer-Sheva, Israel) reviewed and approved the study proposal and the manner in which informed consent was obtained from participants, which was done according to Helsinki human subject guidelines (1983).
|
= 0.05, 1–ß = 0.8, and 6 repeated 24HRs (14). However, because it was suggested (15) that a sample size > 150–200 subjects only slightly increases the accuracy of the CI, and that a sample size < 30 is characterized by a wide CI, we aimed to recruit 30–150 men and women. The FFQ
Preliminary population survey and development of nutritional data base. The Israeli FFQ was developed from an open-ended population survey. A full description of the Negev Nutritional Study (NNS) was given in detail elsewhere (16–18). Briefly, 1173 individuals > 35 y old were recruited using a proportional random cluster sample from voters’ registries of the population in southern Israel. In the NNS, the participants were interviewed in their homes with the use of a modified USDA 24HR questionnaire (19). We developed a food database and data entry system for the 24HR, (20) both based on the U.S. Food Information Analysis System (FIAS) (21). We then modified the USDA food composition database (22,23) and added >2000 Israeli-manufactured foods and 300 representative local recipes (20).
Development of FFQ. The development of the FFQ was described in detail elsewhere (24–26). Briefly, based on the 24HR population survey (NNS), we aggregated conceptually similar foods into 194 groups. Items were entered into stepwise multiple regression analyses (11) according to their specific nutrient content and the frequency of their intake in the population. Foods in the model that accounted for at least 80% of the between-person variability were considered for the final questionnaire. The procedure described was undertaken for each of 27 nutrients. We identified 126 food groups that were the main contributors to the between-person variation of each nutrient. This questionnaire was tested in a pilot study (data not shown).
The final FFQ. The FFQ presents 9 frequency options from "never or less than once a month" to "6 or more a day." It is a semiquantitative questionnaire with a standard portion size provided for each food item. For all items, the questionnaire enables reporting the quantity usually eaten by adding the number of portions eaten in an open column. For example, a person who usually eats 2 slices of bread will insert the number 2 in the open column. The nutrient analysis for each food group was defined by the frequency of intake, by gender and age, of the specific food items in the food groups, as reported in the population-based survey (NNS). Fifteen foods are accompanied by 3 model amounts in a food-pictures booklet. Most of the pictures were chosen from the Epic-Soft Picture courtesy of the International Agency for Research on Cancer, Lyon. Additional pictures of local dishes were chosen from a quantity food-pictures booklet of Palestinian researchers, courtesy of Al-Quds University, Jerusalem.
24HR interviews. Six 24HRs interviews were performed in each home on random workdays. An interview after the weekend randomly asked about either the previous day or previous 2 days. Interviewers were required to have an academic education and to have participated in 40 h of formal training for this study. To prevent behavior bias, the interviewees did not know the day of the interview in advance. To attenuate the within-interviewer variation, each interviewer was allocated a group of respondents with whom they were in contact throughout the year.
The modified USDA (19) 24HR is a standardized 5-pass method consisting of the following: 1) a "Quick List" pass in which the respondent is asked to list everything eaten or drunk the previous day; 2) a "Forgotten Foods" pass in which a standard list of food/beverages, often forgotten, is read to prompt recall; 3) a "Time and Occasion" pass in which the time and name of the eating occasion are collected; 4) a "Detail" pass in which detailed descriptions and portion sizes are collected and the time interval between meals is reviewed to check for additional foods; and 5) the "Final" pass, which provides one last opportunity for the respondent to remember foods consumed. For each food reported, interviewers referred to a standardized food-pictures and models booklet. The recall data were linked to a nutrient database (20).
The 24 h-urine collections and clinical measurements. Participants received detailed written and oral instructions on how and when to collect their urine. They were asked to begin collecting the 2nd urine void of the morning and to include the 1st morning void the following day. During each collection period, participants recorded the time of start and finish and the medications/supplements taken. The volume of the urine sample was recorded after it was received. After the sample was mixed, a 5-mL sample was transferred to a urine test tube. Samples were stored at –20°C until further analysis. Laboratory measurements were performed within 1 d of collection. The participants in this research center were accustomed to routine 24-h urine collection for monitoring absorption of radioactive substances; thus, we did not assess completeness of collection. Urea was measured with a kinetic test with urease and glutamate dehydrogenase (27) in the University-affiliated Medical Center (Soroka Medical Center). Weight and height were measured at the clinic at the time of collection of the first urine sample. The participants were weighed without shoes on a digital scale.
Blood draw.
Blood (20 mL) was drawn in the workplace after participants had fasted overnight and was assayed for -tocopherol, ß-carotene, folic acid, total cholesterol, and other biomarkers. Immediately after blood was drawn, samples were separated into plasma and serum, transferred to the laboratories in iced boxes, and stored at –70°C until further analysis. Serum -tocopherol was extracted into an organic solvent and measured by spectrophotometry at a wavelength of 340 nm (28). Concentration of ß-carotene in a lipid extract of a serum sample was determined spectrophotometrically at a wavelength of 440 nm. Serum folate was assessed by microparticle enzyme immunoassay (Abbott AxSYM System), and lipid profiles were determined by an Eppendorf EPOS 5060 analyzer. The clinical biochemistry laboratory of the University-affiliated Medical Center (Soroka Medical Center) provides most of the biochemistry diagnostic tests for the region (
5 million tests/y). In addition to the controls that accompanied each assay, the quality control (QC) systems used by the laboratory are 1) daily QC system of DADE (USA) and 2) random-sample QC of NEQAS (UK) twice a month. Most of the tests are carried out on a Hitachi 747 autoanalyzer. This apparatus is calibrated on a daily basis with Boehinger’s calibrator for automated systems.
Statistical analyses.
For the FFQ, intakes of fruits and vegetables were adjusted for seasonality where required by considering availability during the year. We used natural log transformation to improve the normal distribution of dietary components, because the values were skewed to the right. The values were adjusted for energy by the residual method (29). We deattenuated (30) the correlation coefficients by multiplying them by the factor [1+(
2[infi]w/2[infi]b)/n]1/2, (31) where n is the number of repeated questionnaires, 2[infi]w is the intraindividual variance, and 2[infi]b is the interindividual variance between questionnaires. We calculated the Pearson correlation coefficient to assess the reliability of the FFQs and their relative validity compared with the other methods. We defined the square roots of the FFQs’ correlations as the estimate for the upper limit of the validity coefficient (32). We controlled for serum cholesterol when calculating -tocopherol and ß-carotene correlations (33). We calculated total protein consumption from the 24-h urine collection with the following equation suggested by Isaksson (34):
 |
The VCs of the dietary intake to the "true intake" (T) were evaluated by the "method of triads" (8,9), as follows, where Q is the questionnaire, R is the reference method, M is the biochemical biomarker, 
or r indicates the correlation estimates, and 
2 indicates the proportion of variation that is related to true intake (35).
 |
 |
 |
All of the analyses were adjusted for gender. When calculating correlations between the dietary questionnaires, we used dietary intake and not vitamin supplements to prevent overestimation of the correlations driven by the vitamin supplements.
|
RESULTS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
|
-tocopherol. The correlations with urinary nitrogen were 0.30 for the 24HR and 0.34 for the FFQ. Correlations with serum ß-carotene with supplements were 0.39 for the 24HR and 0.42 for the FFQ. Serum folic acid with supplements had the highest correlation with the questionnaires (r = 0.41 for 24HR and r = 0.47 for FFQ). In our study population, analysis including vitamin supplements did not appreciably improve the correlations of the biomarkers to the questionnaires.
|
-tocopherol to "true intake," as estimated by "the method of triads," are presented in Figure 1. For protein, ß-carotene, and folic acid, the VCs of FFQ to "true intake" (0.77, 0.65, and 0.72, respectively) were higher than the VCs of 24HR to "true intake" (0.68, 0.60, and 0.39, respectively). However, the performance of the FFQ was inferior to that of the 24HR for estimating -tocopherol (VCs = 0.56 and 0.97, respectively). As for the biomarkers, the VCs of serum ß-carotene and folic acid (0.65 and 0.65) to "true intake" were higher than those of urinary nitrogen and -tocopherol (0.44 and 0.34, respectively). Repeating the entire analysis for the subgroup (70%) who did not take any vitamin supplements did not change the correlations appreciably (Table 7).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
A previous study (13) suggested that a sample size > 120 may be necessary to stabilize the variances in this kind of validation study. Our sample size of 161 participants, using six 24HRs, 2 blood draws, and three 24-h urine collections, allowed for reasonable precision in estimating the attenuation factors for diet-disease RRs. Our study was performed among a highly educated population with few dropouts. However, it is possible that educated participants could overestimate the performance of the FFQ.
An assumption of the attenuation-factor approach is that measurement errors from the FFQ and 24HR are independent relative to a true gold standard. We cannot disregard the potential for correlated errors between the FFQ and 24HRs; therefore, the VCs reported in this study could have been overestimated and might be considered as upper limits of the true VC.
In our study, the FFQ was reproducible, as shown by a range in Pearson’s partial correlation coefficients of 0.43 (folic acid) to 0.67 (energy intake). Energy adjustment attenuated the crude correlations because of the adjustments for the correlated errors. The square root of the reliability is suggested to estimate the upper limit of the VC (32). In our study, the square root ranged from 0.67 to 0.81. With the exception of folic acid, with an estimated VC of FFQ (r = 0.72) that exceeded the square of the reliability (r = 0.67), the estimated upper limit of the VC was higher than or similar to (e.g., protein, r = 0.77) the calculated VC.
The correlation coefficients between nutrients obtained by the FFQ and those obtained by the repeated 24HRs, which ranged from 0.45 (folic acid) to 0.67 (dietary fiber), are similar to those noted between FFQs and 24HRs in other studies (36–41). The most highly correlated nutrients were dietary fiber and vitamin E (even without supplements), which are likely to be concentrated in specific foods. Consistent with other established FFQs, (42) low correlations for energy indicate that the FFQ is better able to discriminate within the population than to estimate their absolute energy intake. Biochemical markers constitute reference measures that have different sources of errors than nutritional questionnaires (43). In comparing the performance of the 2 dietary instruments, we found that the correlation between the FFQ and urinary nitrogen was only slightly higher than that of the mean of 6 repeated 24HRs (r = 0.34 vs. 0.30, P < 0.001 for both), serum ß-carotene (r = 0.42 vs. 0.39, with supplements, P < 0.001 for both), and serum folic acid (r = 0.47 vs. 0.40, with supplements, P < 0.001 for both). This suggests a robust, uniform association between those biomarkers and the 2 dietary assessment methods.
Blood -tocopherol was poorly correlated with the FFQ (0.19, P < 0.05), as also found in a previous study (r = 0.06) (13), and more highly correlated with the 24HR (0.33, P < 0.001). Given the relatively high correlation between vitamin E reported by the 2 dietary assessment methods, we might assume that long-term intake of vitamin E may not be reflected well in blood, and validation against -tocopherol from adipose tissue would have been more appropriate. Thus, if the biomarker is not thought to be a meaningful measure of nutrient status, its application by the method of triads is questionable.
The range of correlation between FFQ and urinary nitrogen (36–41,44) is between 0.2 and 0.4. Others showed similar correlations for plasma ß-carotene with FFQ (13,37,45) as well as correlations < 0.25 (39,46,47).
The "method of triads" also allows estimation of the validity of the biochemical measurements. Measurements of serum ß-carotene (r = 0.67) and serum folic acid (r = 0.65) appeared to have higher VCs than measurements of urinary nitrogen (r = 0.44) and serum -tocopherol (r = 0.34). Consistent with a previous study, (13) the VCs of the biomarkers were not higher than those of the FFQ, suggesting that biomarkers may not perform better than the FFQ. However, this assumption should be carefully considered because the validity of the biomarker data with the dietary measures may not be compared directly due to the different sources of errors.
Published data on the estimation of the attenuation factor by this method are sparse. However, available results vary due to the errors associated with differences in sample sizes, populations, type of questionnaires, or reference methods. For instance, in previous studies that applied the "method of triads" and used serum as a biomarker for ß-carotene and dietary recall as a reference dietary assessment method, the VCs for the FFQ were 0.44 (n = 61) (9), 0.76 (n = 120) (13), and 0.39 (n = 87) (13). In our study, the VC was 0.67 (n = 161). This suggests that specific calibration studies should be applied for specific populations and specific dietary assessment tools and nutrients. Alternatively, a wide range in the attenuation factor should be considered.
The DEARR study showed that the newly developed FFQ is a valid and reproducible instrument for assessing dietary intake. Furthermore, the VCs suggest that biomarkers could be used as estimates for ß-carotene and folic acid rather than for -tocopherol and protein. However, in general, the FFQ appears to perform better than the 24HRs and the biomarkers in estimating "true intake." The VCs obtained can be used for future adjustment of diet-disease RR estimates in this population.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Abbreviations used: DEARR, Dietary Evaluation and Attenuation of Relative Risk; FIAS, Food Information Analysis System; NNS, Negev Nutrition Study; QC, quality control; RR, relative risk; VC, validity coefficient; 24HRs, 24-h recalls.
Manuscript received 9 August 2004. Initial review completed 7 September 2004. Revision accepted 2 December 2004.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. Subar, A. F., Kipnis, V., Troiano, R. P., Midthune, D., Schoeller, D. A., Bingham, S., Sharbaugh, C. O., Trabulsi, J. & Runswick, S., et al (2003) Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN Study. Am. J. Epidemiol. 158:1-13.
2. Kipnis, V., Subar, A. F., Midthune, D., Freedman, L. S., Ballard-Barbash, R., Troiano, R. P., Bingham, S., Schoeller, D. A., Schatzkim, A. & Carroll, R. J. (2003) Structure of dietary measurement error: results of the OPEN biomarker study. Am. J. Epidemiol. 158:14-21.
3. Willett, W. C. (2003) Invited commentary: OPEN questions. Am. J. Epidemiol. 158:22-24.
4. Rosner, B., Willett, W. C. & Spiegelman, D. (1989) Correction of logistic regression relative risk estimates and confidence intervals for systematic within-person measurement error. Stat. Med. 8:1051-1069.[Medline]
5. Kaaks, R. & Riboli, E. (1997) Validation and calibration of dietary intake measurements in the EPIC project: methodological considerations. Int. J. Epidemiol. 26:S15-S25.
6. Kipnis, V., Carroll, R. J., Freedman, L. S. & Li, L. (1999) Implications of a new dietary measurement error model for estimation of relative risk: application to four calibration studies. Am. J. Epidemiol. 150:642-651.
7. Wacholder, S., Armstrong, B. & Hartge, P. (1993) Validation studies using an alloyed gold standard. Am. J. Epidemiol. 137:1251-1258.
8. Kaaks, R. J. (1997) Biochemical markers as additional measurements in studies of the accuracy of dietary questionnaire measurements: conceptual issues. Am. J. Clin. Nutr. 65:1232S-1239S.
9. Ocke, M. C. & Kaaks, R. (1997) Biochemical markers as an additional measurements in dietary validity studies: application of the method of triads with examples from the European Prospective Investigation into Cancer and Nutrition. Am. J. Clin. Nutr. 65:1240S-1245S.
10. Andersen, L. F., Solvoll, K., Johansson, L. R., Salminen, I., Aro, A. & Drevon, C. A. (1999) Evaluation of a food frequency questionnaire with weighed records, fatty acids, and alpha-tocopherol in adipose tissue and serum. Am. J. Epidemiol. 150:75-87.
11. Willett, W. C., Sampson, L., Stampfer, M. J., Rosner, B., Bain, C., Witschi, J., Hennekens, C. H. & Speizer, F. E. (1985) Reproducibility and validity of a semiquantitative food frequency questionnaire. Am. J. Epidemiol. 122:51-65.
12. Daures, J. P., Gerber, M., Scali, J., Astre, C., Bonifacj, C. & Kaaks, R. (2000) Validation of a food- frequency questionnaire using multiple-day records and biochemical markers: application of the triads method. J. Epidemiol. Biostat. 5:109-115.[Medline]
13. Kabagambe, E. K., Baylin, A., Allan, D. A., Siles, X., Spiegelman, D. & Campos, H. (2001) Application of the method of triads to evaluate the performance of food frequency questionnaires and biomarkers as indicators of long -term dietary intake. Am. J. Epidemiol. 154:1126-1135.
14. Rosner, B. (2000) Fundamentals of Biostatistics 2000:463 Harvard University Press Cambridge, MA.
15. Willett, W. C. (1998) Nutritional Epidemiology 1998:33 Oxford University Press Cambridge, MA.
16. Shahar, D., Shai, I., Vardi, H. & Fraser, D. (2003) Dietary intake and eating patterns of elderly people in Israel: who is at nutritional risk?. Eur. J. Clin. Nutr. 57:18-25.[Medline]
17. Bilenko, N., Shahar, D., Shai, I., Weitzman, S. & Fraser, D. (2003) Prevalence and characteristics of myocardial infarction, diabetes and hypertension in the adult Jewish population: results from the Negev Nutritional Study. Harefuah 42:17-21.
18. Fraser, D., Shahar, D., Shai, I., Vardi, H. & Bilenko, N. (2000) Negev nutritional studies: nutritional deficiencies in young and elderly populations. Public Health Rev. 28:31-46.[Medline]
19. Conway, J. M., Ingwersen, L. A., Vinyard, B. T. & Moshfegh, A. J. (2003) Effectiveness of the US Department of Agriculture 5-step multiple-pass method in assessing food intake in obese and nonobese women. Am. J. Clin. Nutr. 77:1171-1178.
20. Shai, I., Vardi, H., Shahar, R. D., Azrad, A. B. & Fraser, D. (2003) Adaptation of international nutrition databases and data entry system tools to a specific population. Public Health Nutr. 6:401-406.[Medline]
21. University of Texas Health Science Center at Houston, School of Public Health (1996) USDA Human Nutrition Information Service, Food Intake Analysis System ver. 3 1996 Washington, DC.
22. U.S. Department of Agriculture, Agricultural Research Service (1999) Composition of Foods, SR 12 1999 USDA Beltsville, MD.
23. U.S. Department of Agriculture, Agriculture Research Service (1976–1987) Composition of Foods Raw. Processed and Prepared. Agriculture Handbook No. 8–1-8–15 1976–1987 U.S. Department of Agriculture Washington, DC.
24. Shahar, D. R., Shai, I., Vardi, H., Brener-Azrad, A. & Fraser, D. (2003) Development of a semi-quantitative Food Frequency Questionnaire (FFQ) to assess dietary intake of multiethnic populations. Eur. J. Epidemiol. 18:855-861.[Medline]
25. Shai, I., Shahar, D. R., Vardi, H. & Fraser, D. (2004) Selection of food items for inclusion in a developed Food Frequency Questionnaire. Public Health Nutr. 7:745-749.[Medline]
26. Shahar, D., Fraser, D., Shai, I. & Vardi, H. (2003) Development of a food frequency questionnaire (FFQ) for an elderly population based on a population survey. J. Nutr. 133:3625-3629.
27. Sampson, E. J., Baird, M. A., Burtis, C. A., Smith, E. M., Witte, C. L. & Bayse, D. D. (1980) A coupled-enzyme equilibrium method for measuring urea in serum: optimization and evaluation of the AACC study group on urea candidate reference method. Clin. Chem. 26:816-826.
28. Vatassery, G. T., Johnson, G. J. & Krezowski, A. M. (1983) Changes in vitamin E concentrations in human plasma and platelets with age. J. Am. Coll. Nutr. 23:69-75.
29. Willett, W. C. & Stampfer, M. J. (1986) Total energy intake: implication for epidemiological analyses. Am. J. Epidemiol. 124:17-27.
30. Rosner, B. & Willett, W. C. (1988) Interval estimates for correlation coefficients corrected for within-person variation: implication for study design and hypothesis testing. Am. J. Epidemiol. 127:377-388.
31. Liu, K., Stamler, J., Dyer, A., McKeever, J. & McKeever, P. (1978) Statistical methods to assess and minimize the role of intra-individual variability in obscuring the relationship between dietary lipids and serum cholesterol. J. Chron. Dis. 31:399-418.[Medline]
32. Armstrong, B., White, E. & Saracci, R. (1992) Principles of Exposure Measurement in Epidemiology 1992:78-114 Oxford Medical Publications Oxford, UK.
33. Stryker, W. S., Kaplan, L. A., Stein, E. A., Stampfer, M. J., Sober, A. & Willett, W. C. (1988) The relation of diet, cigarette smoking, and alcohol consumption to plasma beta-carotene and alpha-tocopherol levels. Am. J. Epidemiol. 127:283-296.
34. Isaksson, B. (1980) Urinary nitrogen output as a validity test in dietary surveys. [letter]Am. J. Clin. Nutr. 33:4-5.
35. Dunn, G. (1989) Design and Analysis of Reliability Studies: The Statistical Evaluation of Measurement Error 1989 Oxford University Press New York, NY.
36. Carlos, A. G. (1997) Relative validity and reproducibility of a diet history questionnaire in Spain. Int. J. Epidemiol. 26:s91-s99.
37. Stefanie, B. T., Ina, H. & Heiner, B. (1997) Reproducibility and relative validity of food group intake in a FFQ developed for the German part of the EPIC project. Am. J. Epidemiol. 26:S59-S70.
38. Ocke, M. C. & Bueno-De Mesquita, H. B. (1997) The Dutch EPIC FFQ. l. Description of the questionnaire, and relative validity and reproducibility for food groups. Am. J. Epidemiol. 26:S37-S48.
39. Klea, K., Eric, B. R. & Charalambos, G. (1997) Reproducibility and relative validity of an extensive semi-quantitative FFQ using dietary records and biochemical markers among Greek schoolteachers. Int. J. Epidemiol. 26:S118-S127.
40. Paola, P. & Fabrizio, F. (1997) Relative validity and reproducibility of FFQ for USA in the Italian EPIC centers. Int. J. Epidemiol. 26:S152-S159.
41. Marti, J.V.L. & Francois, L. (1997) Relative validity and reproducibility of a French dietary history questionnaire. Int. J. Epidemiol. 26:S128-S136.
42. Subar, A. F., Thompson, F. E., Kipnis, V., Midthune, D., Hurwitz, P., McNutt, S., McIntosh, A. & Rosenfeld, S. (2001) Comparative validation of the Block, Willett, and National Cancer Institute FFQs. Am. J. Epidemiol. 154:1089-1099.
43. Willett, W. C., Stampfer, M. J., Underwood, B. A., Speizer, F. E., Rosner, B. & Hennekens, C. H. (1983) Validation of a dietary questionnaire with plasma carotenoid and alpha-tocopherol levels. Am. J. Clin. Nutr. 38:631-639.
44. McKeown, N. M., Day, N. E., Welch, A. A., Runswick, S. A., Luben, R. N., Mulligan, A. A., McTaggart, A. & Bingham, S. A. (2001) Use of biological markers to validate self-reported dietary intake in a random sample of the European Prospective Investigation into Cancer United Kingdom Norfolk cohort. Am. J. Clin. Nutr. 74:188-196.
45. Coates, R. J., Eley, W., Block, G., Gunter, E. W., Sowell, A. L., Grossman, C. & Greenberg, R. S. (1991) An evaluation of a FFQ for assessing dietary intake of specific carotenoid and vitamin E among low-income black women. Am. J. Epidemiol. 134:658-670.
46. Roidt, L., White, E., Goodman, G., Wahl, P. W., Omenn, G. S., Rollins, B. & Karbeck, J. M. (1988) Association of food frequency questionnaire estimates of vitamin A intake with serum vitamin A levels. Am. J. Epidemiol. 128:645-654.
47. Liu, T., Wilson, N. P., Craig, C. B., Tamura, T., Soong, S., Sauerlich, H. E., Cole, P. & Butterworth, C. E., Jr (1992) Evaluation of three nutritional assessment methods in a group of women. Epidemiology 3:496-502.[Medline]
This article has been cited by other articles:
|
|
 |
|
  I. Shai, D. Schwarzfuchs, Y. Henkin, D. R. Shahar, S. Witkow, I. Greenberg, R. Golan, D. Fraser, A. Bolotin, H. Vardi, et al. Weight Loss with a Low-Carbohydrate, Mediterranean, or Low-Fat Diet N. Engl. J. Med., July 17, 2008; 359(3): 229 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Bingham, R. Luben, A. Welch, Y. L. Low, K. T. Khaw, N. Wareham, and N. Day Associations between dietary methods and biomarkers, and between fruits and vegetables and risk of ischaemic heart disease, in the EPIC Norfolk Cohort Study Int. J. Epidemiol., June 25, 2008; (2008) dyn111v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Shahar, R. Abel, A. Elhayany, H. Vardi, and D. Fraser Does Dairy Calcium Intake Enhance Weight Loss Among Overweight Diabetic Patients? Diabetes Care, March 1, 2007; 30(3): 485 - 489. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. B. Dixon, A. F. Subar, L. Wideroff, F. E. Thompson, L. L. Kahle, and N. Potischman Carotenoid and Tocopherol Estimates from the NCI Diet History Questionnaire Are Valid Compared with Multiple Recalls and Serum Biomarkers J. Nutr., December 1, 2006; 136(12): 3054 - 3061. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. C. Willett and F. B. Hu Not the Time to Abandon the Food Frequency Questionnaire: Point Cancer Epidemiol. Biomarkers Prev., October 1, 2006; 15(10): 1757 - 1758. [Full Text] [PDF]
|
|
|
|
 |
|
 L. Natarajan, S. W. Flatt, X. Sun, A. C. Gamst, J. M. Major, C. L. Rock, W. Al-Delaimy, C. A. Thomson, V. A. Newman, J. P. Pierce, et al. Validity and Systematic Error in Measuring Carotenoid Consumption with Dietary Self-report Instruments Am. J. Epidemiol., April 15, 2006; 163(8): 770 - 778. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
