![[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}
|
|
|
|
|
||
3 Applied Research Program, Division of Cancer Control and Population Sciences, and 4 Biometry Program, Division of Cancer Prevention, National Cancer Institute, Bethesda, MD 20892-7344; 5 Departments of Medicine, Clinical and Social Sciences in Psychology, Psychiatry, University of Rochester, Rochester, NY 14642; 6 Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Bethesda, MD 20892-7344; 7 Department of Epidemiology and Biostatistics, Arnold School of Public Health and Cancer Prevention and Control Program, University of South Carolina, Columbia, SC 29208; 8 Health Behavior and Health Education, School of Public Health, University of Michigan, Ann Arbor, MI 48109-2029; 9 Oregon Research Institute, Eugene, OR 97403; 10 Department of Nutrition and Food Sciences, University of Rhode Island, Kingston, RI 02881; and 11 Department of Nutrition and Department of Society, Human Development and Health, Harvard School of Public Health, Cambridge, MA 02115
* To whom correspondence should be addressed. E-mail: thompsof{at}mail.nih.gov.
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionSubjects and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
Introduction |
|---|
|
TOPABSTRACTIntroductionSubjects and MethodsResultsDiscussionLITERATURE CITED |
|---|
Intervention researchers require precise, reproducible instruments to measure fat intake. The interviewer-administered 24-h dietary recall (24HR)12 provides the most accurate and complete self-reported information about the individual's diet for a given day (13). However, because this method currently requires highly trained interviewers and imposes a heavy investigator burden with regard to administration, coding, and data processing, it is prohibitively expensive and therefore not feasible for many research applications. In some situations, self-administered FFQ, often optically scannable for inexpensive data entry, are used. However, comprehensive FFQ consisting of >100 items often require up to 1 h to complete and, if self-administered, require a high level of literacy. Furthermore, FFQ have been found to have substantial measurement error (14–17).
Various shorter tools that measure a limited number of dietary factors rather than the entire diet have been developed (13). Although they are more feasible than longer instruments to administer, they are limited in the amount of information they capture. Numerous short instruments have been developed to assess fat intake in U.S. populations. Most rank individuals based on their fat intake but do not attempt to estimate the total absolute amounts of fat consumed (18–24). Furthermore, because dietary guidance is given in terms of percentage energy from fat and saturated fat, instruments that estimate these parameters are desirable. Absolute fat intake and percentage energy from fat measure 2 different constructs. For example, those who eat a lot may have high fat intakes but also relatively low percentage energy from fat. Two instruments have been developed that quantify individual fat intake as a percentage of total energy (25,26). The Behavior Change Consortium (BCC), an NIH-funded set of intervention trials, chose to use the NCI percentage of energy from fat short instrument (PFat) (26).
The purpose of this article is to evaluate the performance of the PFat in a set of intervention studies at baseline. We compare estimates from the PFat instrument to those from multiple 24HR, controlling for within-person variability in the 24HR. In addition, we examine factors that may moderate the screener's performance. Results of analyses evaluating the ability of the screener to measure change caused by the various interventions are presented in another article in this supplement (27).
|
Subjects and Methods |
|---|
|
TOPABSTRACTIntroductionSubjects and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
The PFat consists of 16 questions that ask about usual consumption of foods over the past year (28). The foods on the screener were identified as those that best explained variability in percentage energy from fat in a nationally representative sample of adults. Foods could be positive predictors of percentage energy from fat, e.g., bacon, French fries, or could be negative predictors of percentage energy from fat, e.g., fruit, cold cereal, skim milk. Some major sources of fat in the U.S. population, e.g., beef, fried chicken, are not major predictors of variability in percentage energy from fat and so are not included on the screener. Estimates of percentage of energy from fat are computed from respondent-reported frequency responses, assigned externally derived gender- and age-specific portion sizes, and gender-specific regression coefficients using USDA's 1994–1996 Continuing Survey of Food Intakes by Individuals. Details of the tool's development, scoring, and testing are given by Thompson et al. (26). The PFat is available electronically (29).
The total baseline sample consisted of 1474 participants. Of these, 524 (36%) completed both the screener and at least 1 recall, 855 (58%) completed only the PFat, and 95 (6%) completed only 24HR. In this article, we include for analysis and presentation data from only those 524 participants who completed both the PFat and 24HR. Of these, 74% had 3 24HR; 19% had 2 24HR; and 7% had a single 24HR.
Analytical procedures. Validity is defined as the concurrence between measured and true exposure. Although true usual dietary intake in free-living populations is impossible to measure (30), its distribution in the population can be estimated, as can the relation between true and screener-reported intake, by use of an appropriate reference instrument and statistical methods. The reference instrument used in the BCC study is multiple 24HR and is analyzed using a latent variable measurement error model, described by Freedman et al. (31).
Measurement error is defined as the difference between reported and true exposure and is conceptualized as being composed of systematic bias and within-person random error. The systematic bias may be related to true intake or to other characteristics such as age or BMI, whereas the within-person error has a mean of 0 and is unrelated to any other measurements. True intake is modeled as a latent variable using information from repeated measures of the reference instrument. The screener is assumed to have systematic bias and within-person random error, whereas the reference instrument is assumed to have within-person random error but no systematic bias. The model can incorporate covariates such as age or BMI to allow for systematic biases other than those related to true intake.
Objective biomarker measures are available for energy and protein intake (14–17) but not for fat and percentage energy from fat (32). Self-report measures of diet are associated with error from a variety of sources. The 24HR is considered one of the better self-report methods available, not only because of the detailed description of the diet obtained but also because lack of literacy among potential respondents does not adversely impact the quality of information obtained, and because, if obtained without prior notice, there is no reactive effect (13). However, even with its advantages, the 24HR has been found in biomarker studies to contain individual-level bias for some nutrients, with bias toward underreporting of energy (14,16,17,33,34). In the BCC, we used multiple 24HR as the criterion gold standard, albeit an "alloyed" gold standard.
Our analytical objectives in this intervention context were several. First, we evaluated the ability of the PFat to estimate the mean and distribution of percentage energy from fat intake in the population and to adequately characterize the baseline intake for comparison to postintervention intake. Second, we evaluated the ability of the PFat to estimate intake of an individual and to rank an individual's intake within the population, useful to determine eligibility for entry into the study and/or later assignment to the appropriate treatment group. Third, we examined whether other factors moderated the relation between the screener and true intake. Because preliminary analyses revealed significant site and gender differences in all parameters, we performed all analyses presented here stratified by site and gender.
We assessed the ability of the PFat to estimate mean intake in the population by comparing means of individuals' percentage energy from fat values from multiple recalls to that from their screener and used a paired t-test to test for differences, using P < 0.05. To see how well the PFat estimates other characteristics of the distribution of intake, we compared prevalence estimates from the screener to true prevalence as estimated from multiple 24HR in the measurement error model. Prevalence is the proportion of individuals in the population above or below a specified level of intake. The deattenuated Pearson correlation coefficient and its standard error also are estimated from the model and are used for evaluation of the PFat's ability to rank individuals accurately. The squared correlation coefficient (R2) measures the proportion of variation in true intake explained by reported intake; a correlation of 0.5 indicates that 
25% of the variation in true intake is explained by the PFat screener.
We assessed the ability of the PFat to estimate individual intake by estimating the screener's positive and negative predictive values at intake levels at or <30% energy from fat [recommended in Healthy People 2010 (1)] and <35% energy from fat [recommended by Dietary Guidelines for Americans (2) and Institute of Medicine (3)]. The positive predictive value of a screening tool represents the proportion of subjects selected into the study who are truly eligible, whereas the negative predictive value represents the proportion of subjects excluded from the study who were truly ineligible. For an instrument to be useful as a screening tool, the positive predictive value should be substantially larger than the true prevalence in the population. Positive and negative predictive values were estimated using the measurement error model.
The scoring algorithm for the screener was developed in an external data set (Continuing Survey of Food Intakes by Individuals, 1994–1996) and uses a regression model to estimate the conditional expectation of true intake given the screener responses. Because the screener estimate derives from a conditional expectation, mathematically, the screener estimate should have about the same mean as, but smaller variance than, true intake (35) and so, without any adjustment, will provide poor estimates of prevalence and have poor sensitivity. Thompson et al. (35) suggest multiplying the screener by a variance adjustment factor so that its variance more closely approximates the variance of true intake in the population. The formula for the variance-adjusted screener is:
 |
In the NIH-AARP Diet and Health Study, the estimated variance adjustment factor for the PFat screener was 2.0 for men and 1.7 for women (26). We applied these external variance adjustment factors to the PFat screener in the BCC samples when estimating prevalence and positive predictive value.
We examined agreement between the screener and the 24HR by treatment group (control, fat as secondary intervention, fat as primary intervention) to confirm effective randomization of participants to treatment group. In addition, we examined whether other factors moderated the agreement between the screener and the 24HR. These included individual demographic variables: years of age (18–39, 40–59, 60 or older), educational status (less than high school, high school only, more than high school); and standard defined categories of BMI [weight (kg)/height (m)2]—normal weight (<18.5–24.9) overweight (25–29.9), and obese (
30) (36). We created 2 variables to characterize the level of agreement in mean estimates between the 24HR and screener: the difference score, defined as 24HR – screener, and the ratio score, defined as 24HR/screener. The difference score would characterize the amount and direction of bias, whereas the ratio score would characterize the proportional bias. Because both scores appeared to be approximately normally distributed, analyses of variance methods were used to test whether mean scores differed among different subgroups using P < 0.05. Extreme scores, as defined below, were excluded from these analyses. Because of cross-site differences in the distribution of potential moderating variables, sample sizes were not sufficient to test all variables in all sites. (Small sample size and consequently unstable model estimates also precluded examination of differences in correlation coefficients across site/genders.)
In our analysis, the measurement error model we used assumes all variables are normally distributed. Because percentage fat was in this study approximately normally distributed for both the PFat and 24HR, no transformation to normality was needed. Before analysis, we excluded extreme values of percentage energy from fat from each instrument to avoid their undue influence. For each gender, values >3 interquartile ranges below quartile 1 or 3 interquartile ranges above quartile 3 of that variable's distribution were excluded (for each of 3 d of 24HR, percentage energy from fat values were –10.3 to –17.3 and 73.4 to 80.2 for males and –15.4 to –17.9 and 80.0 to 80.8 for females; for PFat, percentage energy from fat values were 15.2 and 45.8 for males and 9.6 and 50.9 for females). This procedure was followed for each day of dietary recall and for the screener. Under these criteria, no values were excluded for the 24HR; 1 value (57.4) was excluded for the screener. Subsequent to fitting the measurement error model, diagnostic statistics were used to identify influential observations. Three observations (2 women in Emory and 1 woman in ROC) were identified that had a significant impact on the estimate of the correlation coefficient. Thus, deattenuated correlations are presented excluding these 3 subjects.
|
Results |
|---|
|
TOPABSTRACTIntroductionSubjects and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
Estimated prevalence of being at or below 30% energy from fat and <35% energy from fat for true intake and screener are presented in Table 3. Prevalence estimates at or below 30% energy from fat were identical for true intake and screener among URI men but were quite different among Emory women and ROC men. The positive predictive values were at least 10% larger than the true prevalence in the population for all sites and at least 24% larger for all sites except Emory. The negative predictive values were 0.61 or higher for all sites. At 35% energy from fat, prevalence estimates between true intake and screener were similar among URI men and women and ROC men but differed greatly for women from HSPH, Emory, and ROC. Also presented in Table 3 are estimates of the percentage energy from fat values where sensitivity and specificity are approximately equal by site and gender. These estimates of optimal screening level vary across sites and genders.
|
Table 4 shows the mean difference scores for the BMI subgroups. (Results for the ratio score are qualitatively similar and are not presented.) Agreement between 24HR and PFat was significantly (P < 0.05) higher in normal-weight than in overweight and obese categories in URI women, and although not statistically significant, followed the same pattern in Emory women and ROC men.
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionSubjects and MethodsResultsDiscussionLITERATURE CITED |
|---|
The NCI PFat instrument used in this BCC Nutrition Working Group validation study has been evaluated in 1 other study. In that study, the PFat was compared with 2 nonconsecutive 24HR in a subsample of 401 men and women participating in the NIH-AARP Diet and Health Study, a prospective cohort study (26). In the study, mean estimated percentage energy from fat intake was significantly lower for the screener than for 24HR in women (28.4 vs. 31.3) but not in men (26). Deattenuated Pearson correlations between the 2 instruments were 0.64 for men and 0.58 for women. The study sample was comprised of older adults (ages 50–71 y), members of AARP (formerly the American Association of Retired Persons); 91% were white, and 72% had higher than high school education. To our knowledge, this BCC study constitutes the first evaluation of the screener in an intervention study and among a diverse group of study samples.
In this study, performance of the PFat varied across sites and genders. There was only limited agreement between screener and 24HR in mean estimated intakes—means for the 2 instruments were statistically significantly different (P < 0.05) or nearly so (0.05 < P < 0.10) from true intake for 4 of 7 site-gender subgroups. Except for men and women at Emory, the magnitude of these statistically significant differences was 2.1 percentage points or less. In the context of an intervention study, the level of error in a screener at baseline is less important than the error in estimating change over time among the same individuals, the area addressed by Williams et al. (27).
Short instruments may be used in dietary intervention studies as screening devices for selection into, or exclusion from, a study. The aim of such screening is to recruit a sample of subjects having, say, true percentage energy from fat >35%; because of misclassification, however, not all subjects selected into the study will be truly eligible. The BCC study indicates that the PFat screener would be only moderately useful in increasing the proportion of truly eligible subjects in a study, at least if one screens at the same level as the desired true level, as we did in this analysis. One could improve the proportion of subjects with true percentage energy from fat >35% by screening at a higher level, say 37%. This would lead to a lower negative predictive value, however, and as a result would require one to screen a larger pool of potential subjects to achieve recruitment goals.
In this study, site- and gender-specific deattenuated correlations between the PFat and 24HR ranged widely, from 0.36 to 0.77. All correlations were positive and statistically significantly different from 0, and the proportion of variance in true percentage energy from fat explained by the screener ranged from 13 to 59%. In the context of an intervention study, it is not enough to show that the screener is correlated with true intake at baseline. One must also establish similar levels of correlation at follow-up and at postintervention, so that any changes in reported intake would be due to actual dietary changes, and not to changes in the validity of the instrument that might occur over time or with the intervention, or both.
If factors that moderate validity could be identified, these could be used to explain differences seen between sites, and could possibly be included with the screener in analytical models to improve overall predictive value. In the BCC, in general, the PFat did not perform as well in women as in men. Some studies have shown higher underreporting of energy and fat among overweight and obese than among normal-weight individuals (39,40). Although there was some indication of such an effect among URI women and possibly ROC men and Emory women, the effect was not evident for all groups examined. Differences in performance were not associated with differences in age, educational status, or smoking status. The lack of moderating effects on validity for a wide range of variables is somewhat surprising. Several factors may have contributed to this. First, the effective sample size for examining these types of moderating effects was quite limited, as these variables did not vary generally within each site, and thus statistical power was low. Second, because a screener instrument is so limited, it may not be sensitive enough to detect small differences in performance among various subgroups. Furthermore, in our analyses of potential effect moderators, we were not able to disentangle design variables such as mode of screener administration that were specific to each site from effects of site itself.
Overall, the PFat did not perform as well in the baseline BCC data presented in the current study as in the NIH-AARP validation study. An important distinction is that this BCC study occurred in the context of an intervention setting, whereas the previous NIH-AARP validation study occurred within a prospective observational study. Participation in an intervention study requires a higher level of commitment than that required in an observational study, which could result in a sample that differs from the overall population in ways that affect measurement of the validity of the instrument in that sample (sometimes called selection bias). For example, intervention participants may be particularly inclined to "talk a good diet"; it is thought that this type of response bias may be more problematic in food frequency-type instruments than in 24HR. However, comparison of the NIH-AARP validation study results to those from URI, the sample most similar demographically to the NIH-AARP sample, reveals very similar agreement in means and correlations. Thus, although the potential for selection bias still exists, it may not be the most important reason for the poorer overall performance of the PFat in the BCC. A more important reason may stem from the diversity of study sites.
The PFat performed somewhat more poorly among Emory men and women than among those in other sites. Although the results for Emory men might result from the small sample size (n = 18), the results for Emory women were based on a relatively large sample (n = 147), thus minimizing the possibility of unstable estimates. The Emory screener was self-administered, which may have led to poorer quality data compared with interviewer-administered instruments. Alternatively, it may be that the relationships between the foods assessed on the screener and percentage energy from fat differ for various subgroups of the population and in particular are different for African Americans in the Emory site. Thus, it may be necessary to develop separate scoring algorithms for specific subpopulations defined by, for example, race/ethnicity or BMI status. One might even need to include particular foods important to estimating percentage energy from fat in specific subpopulations. More research is needed in large representative population samples to model population-specific algorithms. Testing of such population-specific algorithms should be done in both cross-sectional and intervention studies.
In summary, the ability of the PFat to estimate usual intake of percentage energy from fat at baseline in the BCC intervention studies varied across genders and sites. Generally, performance was better among men than women and was somewhat poorer in the site composed of Southern African-Americans. Targeted food lists and/or scoring algorithms may be necessary for certain subgroups of the population. If accurate assessment of diet at baseline (and presumably at follow-up) is essential, a more detailed instrument such as the 24HR, administered multiple times for each participant, is warranted. A self-administered automated 24HR is being developed currently (41) for public use at minimal charge, and it promises to offer researchers a more precise yet affordable option in the near future. If less precise assessment is acceptable, the PFat may be appropriate but should be pretested rigorously in the target population.
|
FOOTNOTES |
|---|
2 Author disclosures: F. E. Thompson, D. Midthune, G. C. Williams, A. L.Yaroch, T. G. Hurley, K. Resnicow, J. R. Hebert, D. J. Toobert, G. W. Greene, K. Peterson, and Linda Nebeling, no conflicts of interest.
12 Abbreviations used: 24HR, 24-h dietary recall; BCC, Behavioral Change Consortium; HSPH, Harvard School of Public Health; NCI, National Cancer Institute; PFat, Percentage Energy from Fat short instrument; ROC, University of Rochester; URI, University of Rhode Island.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTIntroductionSubjects and MethodsResultsDiscussionLITERATURE CITED |
|---|
1. US Department of Health and Human Services. Tracking healthy people 2010. Washington, DC: US Government Printing Office; 2000.
2. US Department of Health and Human Services and US Department of Agriculture. Dietary guidelines for Americans 2005. 6th ed. Washington, DC: US Government Printing Office; 2005.
3. Institute of Medicine. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids. Washington, DC: IOM; 2002.
4. Hebert JR, Patterson RE, Gorfine M, Ebbeling CB, St Jeor ST, Chlebowski RT. Differences between estimated caloric requirements and self-reported caloric intake in the Women's Health Initiative. Ann Epidemiol. 2003;13:629–37.[Medline]
5. DISC Collaborative Research Group. Dietary intervention study in children (DISC) with elevated low-density-lipoprotein cholesterol. Design and baseline characteristics. Ann Epidemiol. 1993;3:393–402.[Medline]
6. Astrup A, Grunwald GK, Melanson EL, Saris WH, Hill JO. The role of low-fat diets in body weight control: a meta-analysis of ad libitum dietary intervention studies. Int J Obes Relat Metab Disord. 2000;24:1545–52.[Medline]
7. Lanza E, Schatzkin A, Daston C, Corle D, Freedman L, Ballard-Barbash R, Caan B, Lance P, Marshall J, et al. Implementation of a 4-y, high-fiber, high-fruit-and-vegetable, low-fat dietary intervention: results of dietary changes in the Polyp Prevention Trial. Am J Clin Nutr. 2001;74:387–401.
8. Barnard ND, Scialli AR, Turner-McGrievy G, Lanou AJ, Glass J. The effects of a low-fat, plant-based dietary intervention on body weight, metabolism, and insulin sensitivity. Am J Med. 2005;118:991–7.[Medline]
9. Howard BV, Van Horn L, Hsia J, Manson JE, Stefanick ML, Wassertheil-Smoller S, Kuller LH, LaCroix AZ, Langer RD, et al. Low-fat dietary pattern and risk of cardiovascular disease: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295:655–66.
10. Beresford SA, Johnson KC, Ritenbaugh C, Lasser NL, Snetselaar LG, Black HR, Anderson GL, Assaf AR, Bassford T, et al. Low-fat dietary pattern and risk of colorectal cancer: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295:643–54.
11. Prentice RL, Caan B, Chlebowski RT, Patterson R, Kuller LH, Ockene JK, Margolis KL, Limacher MC, Manson JE, et al. Low-fat dietary pattern and risk of invasive breast cancer: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295:629–42.
12. Winters BL, Mitchell DC, Smiciklas-Wright H, Grosvenor MB, Liu W, Blackburn GL. Dietary patterns in women treated for breast cancer who successfully reduce fat intake: the Women's Intervention Nutrition Study (WINS). J Am Diet Assoc. 2004;104:551–9.[Medline]
13. Thompson FE, Subar AF. Dietary assessment methodology. In: Coulston AM, Rock CL, Monsen ER, editors. Nutrition in the prevention and treatment of disease. San Diego: Academic Press; 2001.
14. 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 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.
15. Andersen LF, Tomten H, Haggarty P, Lovo A, Hustvedt BE. Validation of energy intake estimated from a food frequency questionnaire: a doubly labelled water study. Eur J Clin Nutr. 2003;57:279–84.[Medline]
16. 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.
17. 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.
18. Block G, Clifford C, Naughton MD, Henderson M, McAdams M. A brief dietary screen for high fat intake. J Nutr Educ. 1989;21:199–207.
19. Kristal AR, Shattuck AL, Henry HJ. Patterns of dietary behavior associated with selecting diets low in fat: reliability and validity of a behavioral approach to dietary assessment. J Am Diet Assoc. 1990;90:214–20.[Medline]
20. Kemppainen T, Rosendahl A, Nuutinen O, Ebeling T, Pietinen P, Uusitupa M. Validation of a short dietary questionnaire and a qualitative fat index for the assessment of fat intake. Eur J Clin Nutr. 1993;47:765–75.[Medline]
21. Dobson AJ, Blijlevens R, Alexander HM, Croce N, Heller RF, Higginbotham N, Pike G, Plotnikoff R, Russell A, et al. Short fat questionnaire: a self-administered measure of fat-intake behaviour. Aust J Public Health. 1993;17:144–9.[Medline]
22. Coates RJ, Serdula MK, Byers T, Mokdad A, Jewell S, Leonard SB, Ritenbaugh C, Newcomb P, Mares-Perlman J, et al. A brief, telephone-administered food frequency questionnaire can be useful for surveillance of dietary fat intakes. J Nutr. 1995;125:1473–83.
23. Retzlaff BM, Dowdy AA, Walden CE, Bovbjerg VE, Knopp RH. The Northwest Lipid Research Clinic Fat Intake Scale: validation and utility. Am J Public Health. 1997;87:181–5.
24. Kris-Etherton P, Eissenstat B, Jaax S, Srinath U, Scott L, Rader J, Pearson T. Validation for MEDFICTS, a dietary assessment instrument for evaluating adherence to total and saturated fat recommendations of the National Cholesterol Education Program Step 1 and Step 2 diets. J Am Diet Assoc. 2001;101:81–6.[Medline]
25. Block G, Gillespie C, Rosenbaum EH, Jenson C. A rapid food screener to assess fat and fruit and vegetable intake. Am J Prev Med. 2000;18:284–8.[Medline]
26. Thompson FE, Midthune D, Subar AF, Kipnis V, Kahle LL, Schatzkin A. Development and evaluation of a short instrument to estimate usual dietary intake of percent energy from fat. J Am Diet Assoc. 2007;107:760–7.[Medline]
27. Williams GC, Hurley TG, Thompson FE, Midthune D, Yaroch AL, Resnicow K, Toobert DJ, Greene GW, Peterson K, Nebeling L, et al. Performance of a short percentage energy from fat tool in measuring change in dietary intervention studies. J Nutr. 2008;138:212S–217S.
28. Yaroch AL, Nebeling L, Thompson FE, Hurley TG, Hebert JR, Toobert DJ, Resnicow K, Greene GW, Elliot DL, et al. Baseline design elements and sample characteristics for seven sites participating in the Nutrition Working Group of the Behavior Change Consortium. J Nutr. 2008;138:185S–192S.
29. Applied Research Program, National Cancer Institute, National Institutes of Health. Percent energy from fat screener. Available from: www.riskfactor.cancer.gov/diet/screeners/fat. Accessed May 8, 2007.
30. Calvert C, Cade J, Barrett JH, Woodhouse A. Using cross-check questions to address the problem of mis-reporting of specific food groups on Food Frequency Questionnaires. UKWCS Steering Group. Eur J Clin Nutr. 1997;51:708–12.[Medline]
31. Freedman LS, Carroll RJ, Wax Y. Estimating the relation between dietary intake obtained from a food frequency questionnaire and true average intake. Am J Epidemiol. 1991;134:310–20.
32. Arab L. Biomarkers of fat and fatty acid intake. J Nutr. 2003;133:925S–32S.
33. Tran KM, Johnson RK, Soultanakis RP, Matthews DE. In-person vs telephone-administered multiple-pass 24-hour recalls in women: validation with doubly labeled water. J Am Diet Assoc. 2000;100:777–83.[Medline]
34. Hebert JR, Ebbeling CB, Matthews CE, Ma Y, Clemow L, Hurley TG, Druker S. Systematic errors in middle-aged women's estimates of energy intake: comparing three self-report measures to total energy expenditure from doubly labeled water. Ann Epidemiol. 2002;12:577–86.[Medline]
35. Thompson FE, Midthune D, Subar AF, McNeel T, Berrigan D, Kipnis V. Dietary intake estimates in the National Health Interview Survey, 2000: Methodology, results, and interpretation. J Am Diet Assoc. 2005;105:352–63.[Medline]
36. National Heart Lung, and Blood Institute in cooperation with the National Institute of Diabetes and Digestive and Kidney Diseases. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. National Institutes of Health: NIH Publication No. 98–4083, 1998.
37. Willett W. Nutritional epidemiology. 2nd ed. New York: Oxford University Press; 1998.
38. Spencer EH, Elon LK, Hertzberg VS, Stein AD, Frank E. Validation of a brief diet survey instrument among medical students. J Am Diet Assoc. 2005;105:802–6.[Medline]
39. Johnson RK, Soultanakis RP, Matthews DE. Literacy and body fatness are associated with underreporting of energy intake in US low-income women using the multiple-pass 24-hour recall: a doubly labeled water study. J Am Diet Assoc. 1998;98:1136–40.[Medline]
40. Poppitt SD, Swann D, Black AE, Prentice AM. Assessment of selective under-reporting of food intake by both obese and non-obese women in a metabolic facility. Int J Obes Relat Metab Disord. 1998;22:303–11.[Medline]
41. Subar AF, Thompson FE, Potischman N, Forsyth BH, Buday R, Richards D, McNutt S, Hull SG, Guenther PM, et al. Formative research of a quick list for an automated self-administered 24-hour dietary recall. J Am Diet Assoc. 2007;107:1002–7.[Medline]
This article has been cited by other articles:
|
|
 |
|
  S. A. A. Beresford, L. M. Klesges, and H. R. H. Rockett Introduction J. Nutr., January 1, 2008; 138(1): 183S - 184S. [Full Text] [PDF]
|
|
|
|
|
|
A. L. Yaroch, L. Nebeling, F. E. Thompson, T. G. Hurley, J. R. Hebert, D. J. Toobert, K. Resnicow, G. W. Greene, G. C. Williams, D. L. Elliot, et al. Baseline Design Elements and Sample Characteristics for Seven Sites Participating in the Nutrition Working Group of the Behavior Change Consortium J. Nutr., January 1, 2008; 138(1): 185S - 192S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E. Davis, K. Resnicow, A. A. Atienza, K. E. Peterson, A. Domas, A. Hunt, T. G. Hurley, A. L. Yaroch, G. W. Greene, T. Goldman Sher, et al. Use of Signal Detection Methodology to Identify Subgroups of Dietary Supplement Use in Diverse Populations J. Nutr., January 1, 2008; 138(1): 205S - 211S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Williams, T. G. Hurley, F. E. Thompson, D. Midthune, A. L. Yaroch, K. Resnicow, D. J. Toobert, G. W. Greene, K. Peterson, L. Nebeling, et al. Performance of a Short Percentage Energy from Fat Tool in Measuring Change in Dietary Intervention Studies J. Nutr., January 1, 2008; 138(1): 212S - 217S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Hebert, T. G. Hurley, K. E. Peterson, K. Resnicow, F. E. Thompson, A. L. Yaroch, M. Ehlers, D. Midthune, G. C. Williams, G. W. Greene, et al. Social Desirability Trait Influences on Self-Reported Dietary Measures among Diverse Participants in a Multicenter Multiple Risk Factor Trial J. Nutr., January 1, 2008; 138(1): 226S - 234S. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
