5
Food-Based Assessment of Dietary Intake

This chapter addresses the question, What food-based dietary assessment methods hold promise for eligibility determination in WIC based on criteria related to either failure to meet Dietary Guidelines (indicated primarily by not meeting Food Guide Pyramid recommendations) or inadequate intake (indicated by falling below nutrient intake cut-off points based on Dietary Reference Intakes)? To answer the question, the committee examined the scientific basis for the potential performance of food-based methods for eligibility determination at the individual level. This examination required consideration of relevant dietary research at the group level. The committee was most interested in reviewing studies of dietary methods designed to assess the usual1 or long-term intakes of individuals and groups, especially those methods that may have the characteristics that meet the criteria for assessing dietary risk described in Chapter 4. To the extent possible, the committee focused on studies conducted with populations served by WIC: women in the childbearing years, children younger than 5 years of age, and low-income women and children from diverse ethnic backgrounds.

The term food-based dietary assessment methods refers to assessment tools used to estimate the usual nutrient or food intake of an individual or a group. Dietary intake is self-reported by individuals (since direct observation of intake by trained observers is impractical), and therefore poses greater challenges than does using anthropometric or biochemical measures for the determination of

1  

Usual intake is defined as the long-run average intake of food, nutrients, or a specific nutrient for an individual (IOM, 2000a).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 57
Dietary Risk Assessment in the WIC Program 5 Food-Based Assessment of Dietary Intake This chapter addresses the question, What food-based dietary assessment methods hold promise for eligibility determination in WIC based on criteria related to either failure to meet Dietary Guidelines (indicated primarily by not meeting Food Guide Pyramid recommendations) or inadequate intake (indicated by falling below nutrient intake cut-off points based on Dietary Reference Intakes)? To answer the question, the committee examined the scientific basis for the potential performance of food-based methods for eligibility determination at the individual level. This examination required consideration of relevant dietary research at the group level. The committee was most interested in reviewing studies of dietary methods designed to assess the usual1 or long-term intakes of individuals and groups, especially those methods that may have the characteristics that meet the criteria for assessing dietary risk described in Chapter 4. To the extent possible, the committee focused on studies conducted with populations served by WIC: women in the childbearing years, children younger than 5 years of age, and low-income women and children from diverse ethnic backgrounds. The term food-based dietary assessment methods refers to assessment tools used to estimate the usual nutrient or food intake of an individual or a group. Dietary intake is self-reported by individuals (since direct observation of intake by trained observers is impractical), and therefore poses greater challenges than does using anthropometric or biochemical measures for the determination of 1   Usual intake is defined as the long-run average intake of food, nutrients, or a specific nutrient for an individual (IOM, 2000a).

OCR for page 57
Dietary Risk Assessment in the WIC Program WIC eligibility. To use a dietary method to assess an individual’s dietary risk of failure to meet Dietary Guidelines or inadequate intake, the method must have acceptable performance characteristics (described in Chapter 4). The committee focused on available dietary tools with regard to their ability to estimate usual intake and their performance characteristics (validity, reliability, measurement error, bias, and misclassification error). The intent was to determine how well the tools could identify an individual’s WIC eligibility status based on the dietary risk of failure to meet Dietary Guidelines or inadequate intake. The committee considered data related to the correct identification of intakes of nutrients, foods, and food groups since elements from any of these three groupings could be used as the indicator on which a criterion could be based. For example, a method to identify failure to meet Dietary Guidelines must be able to identify accurately a person’s usual intake from each of the five basic food groups of the Food Guide Pyramid. This chapter describes (1) the importance of assessing usual intake, (2) commonly used research-quality dietary assessment methods, including their strengths and limitations, (3) methods that compare food intakes with the Dietary Guidelines, and (4) conclusions about food-based methods for eligibility determination. A FOCUS ON USUAL INTAKE As explained below, dietary assessment for the purpose of determining WIC eligibility must be based on long-term intake or the usual pattern of dietary intake, rather than intake reported for a single day or a few days. In the United States and other developed countries, a person’s dietary intake varies substantially from day to day (Basiotis et al., 1987; Carriquiry, 1999; IOM, 2000a; Nelson et al., 1989; Tarasuk, 1996; Tarasuk and Beaton, 1999). This variation introduces random error in estimates of usual intake. Day-to-day variation in intake arises from multiple biologic and environmental influences such as appetite, physical activity, illness, season of the year, holidays, and personal economic conditions. An individual’s intake may become either more erratic or more monotonous when economic constraints are added to other influences on dietary intake. Relationships Among Daily Nutrient Intakes, Usual Intakes, and a Cut-Off Point Figure 5-1 presents distributions of intake for a hypothetical nutrient X that is normally distributed. It depicts the relationship between the distributions of usual intakes of individuals within a population and the distribution of usual intake for that population (solid line P). L marks the cut-off point for determining whether an individual’s usual intake is above or below a specified cut-off

OCR for page 57
Dietary Risk Assessment in the WIC Program FIGURE 5-1 Relationship Between Distributions of Usual Intakes of Nutrient X for Individuals Within a Population (P) and a Generic Cut-Off Level L. SOURCE: Adapted from Yudkin (1996). level L. The individual values reflected in a dotted line represent the day-to-day intakes of an individual that taken together comprise usual intake. On any given day, Individual A and Individual B can have a dietary intake for a specified nutrient that is at, above, or below L. However, Individual A has a long-term average intake (usual intake) below cutpoint L, whereas Individual B has an average or usual intake above cutpoint L. Compared with a set of recalls, a single recall or day of observation would identify many more individuals as falling below L for most nutrients. Therefore, the accurate approximation of an individual’s usual intake requires data collection over many days (Basiotis et al., 1987; Beaton, 1994; IOM, 2000a; Sempos et al., 1993). Identifying Who Falls Above or Below a Cut-Off Point Estimating the proportion of a population group with a nutrient intake above or below L requires the collection of one day of intake data per person in the population plus an independent second day of intake for at least a subsample of the population (Carriquiry, 1999; IOM, 2000a; Nusser et al., 1996). This pro

OCR for page 57
Dietary Risk Assessment in the WIC Program cedure allows for statistical adjustment of the distribution of nutrient intake for the group. That is, with data from 2 days, one can account for the day-to-day variation in intake that is described in the previous section. The statistical methods used account for day-to-day variability of intake in the population and other factors such as day-of-the-week and the skewness of the intake of nutrient X. However, no method based on one or two recalls is available to identify whether an individual’s usual intake would be above or below L. Variability in Food Intake Turning from nutrients to foods, some individuals are relatively consistent in their intake of a few foods (such as low-fat milk or coffee) from day to day, but they may vary widely in their intake of other foods (e.g., corn or water-melon) (Feskanich et al., 1993). Available data suggest that within-person variability is at least as great a problem in estimating an individual’s food intake as it is in estimating an individual’s nutrient intake. In a German study based on 12 diet recalls per person collected over 1 year, the ratio of within-person to between-person variation in food group consumption was greater than 1.0 for nearly all of the 24 food groups included (Bohlscheid-Thomas et al., 1997). The ratio of within-person to between-person variation ranged from 0.6 for spreads to 65.1 for legumes. The high ratios2 reflect large day-to-day within-person variation in the consumption of different foods. In summary, a large body of literature indicates that day-to-day variation in nutrient and food intake is so large in the United States that one or two diet recalls or food records cannot provide accurate information on usual nutrient and food intake for an individual. OVERVIEW OF RESEARCH-QUALITY DIETARY METHODS FOR ESTIMATING FOOD OR NUTRIENT INTAKE A large body of literature addresses the performance of methods developed to assess dietary intakes and conduct research on diet and health. Four methods—diet history, diet recall (typically 24-hours), food record, and food frequency questionnaire (FFQ)—have been widely studied (Bingham, 1987; Dwyer, 1999; IOM, 2000a; Pao and Cypel, 1996; Tarasuk, 1996; Thompson and Byers, 1994). Most studies of dietary data collection methods focus on the ability of a method to estimate nutrient intake accurately—ranging from just one nutrient to a wide array of them. Some studies examine performance with re 2   The within-person variability is an individual’s day-to-day variability in reported intakes (or intraindividual variability or standard deviation within). The between-person variability (or interindividual variability) is the variability in intakes from person to person. A higher ratio of within- to between-person variability means that the variability of the food or nutrient intake is greater within an individual than the variability between individuals.

OCR for page 57
Dietary Risk Assessment in the WIC Program spect to intake of foods or food groups. The findings discussed in this chapter highlight 24-hour diet recalls and food frequencies, since these are the most commonly used dietary methods in the WIC clinic (see Chapter 2).3 General Characteristics The strengths and limitations of available dietary methods have been extensively reviewed elsewhere (Bingham, 1987; Briefel et al., 1992; Dwyer, 1999; Pao and Cypel, 1996; Tarasuk, 1996; Willett, 2000) and are summarized in Table 5-1. Each of the four methods may be used to provide nutrient intake data, food intake data, or both. Table 5-1 also presents major findings that have implications for use of each of the four methods in the WIC program. In addition, after providing descriptive information about the methods, the table presents two major groups of characteristics that are related to the framework described in Chapter 4—performance characteristics and characteristics related to responsiveness to operational constraints in the WIC setting. These characteristics include the resources required to administer the method (WIC staff, time, and facilities such as computer software), and burden and ability of the client to report or record intake accurately. As shown in Table 5-1, the diet history and FFQ methods attempt to estimate the usual intake of individuals over a long period of time, often the past year. The 24-hour diet recall and food record methods reflect intake over 1 day or a few days. As discussed in the previous section, recalls and records are not good measures of an individual’s usual intake unless a number of independent days are observed.4 On average, diet recalls and food records tend to underestimate usual intake—energy intake in particular. On the other hand, FFQs and diet histories tend to overestimate mean energy intakes, depending on the length of the food lists that are used and subjects’ abilities to estimate accurately the frequency and typical portion sizes of foods they consume. Methods Studies Conducted with Low-Income Women and Children Table 5-2 summarizes the few dietary methods studies that have been conducted with low-income pregnant women and young children or in the WIC population. These studies have been primarily aimed at developing or testing the 3   The dietary history method used in the WIC clinic is not necessarily the traditional diet history method, which takes about one to two hours to administer properly. Food records are not often used because of time limitations and difficulties obtaining complete and accurate records. 4   For some nutrients (such as vitamin A) that are highly concentrated in certain foods, or foods that are eaten sporadically, many days or months of intake may be needed to accurately estimate the usual intake of an individual (IOM, 2000a).

OCR for page 57
Dietary Risk Assessment in the WIC Program TABLE 5-1 Comparison of Performance and Operational Constraints of Selected Dietary Assessment Methods in the WIC Setting Criterion or Characteristic Diet History 24-Hour Diet Recall Definition of diet method An interviewer conducts a 1–2 hr interview with a respondent (or proxy for a child) to ask usual meal patterns, food intake, and other information related to diet Typically includes two or more diet methods (food frequency questionnaire [FFQ], 24-hr diet recall, 3-d food record, or questionnaire on diet behaviors), but a standardized method is not available An interviewer asks the respondent (or proxy for a child) to recall all foods and beverages consumed yesterday (for a 24-hr period such as midnight to midnight); food descriptions and amounts for each food are recalled; amounts are estimated using portion size measurement aids Ability to estimate usual food or nutrient intake (an individual’s average intake over a long period of time) Yields a more representative pattern of usual intakes in the past than other methods; generally designed to assess total diet Tends to overestimate nutrient intakes compared with diet recall and food record Provides information on the frequency and types of foods typically eaten, preparations, and detailed descriptions of foods Quantification of intake imprecise due to poor recall or use of standard portion sizes Reflects a single day’s intake rather than usual intake (not a valid estimate of an individual’s usual intake); several or many days over a defined time period are required to estimate usual nutrient intake Number of days needed to estimate usual intake depends on desired precision of estimate Provides quantitative estimates of foods and nutrients Tends to underestimate energy intake Provides information on food details and food preparation methods for single days of intake

OCR for page 57
Dietary Risk Assessment in the WIC Program Food Record Food Frequency Questionnaire (FFQ) The respondent (or proxy for a child) records all foods and beverages, food descriptions including preparation and ingredients, and food amounts for a specified period of time, typically recorded as consecutive 3, 7, or 14 d, but can also be nonconsecutive daily records over a period of time The respondent (or a proxy for a child) completes a questionnaire that asks about the frequency of consumption of foods and beverages over a specified period (1 mo, 3 mo, or 1 yr); may or may not ask about portion sizes Usually self-administered One-day records kept intermittently over a year may reflect an individual’s usual intake Multiple records may be required to estimate usual nutrient intake Provides quantitative estimates of foods and nutrients Tends to underestimate energy intake Foods eaten away from home are less accurately described than those eaten at home Useful to assess qualitative intake and dietary patterns Designed to estimate usual intake of foods; semiquantitative methods are used to estimate nutrients from food frequency information; useful for estimating foods that are consumed frequently, infrequently, or never Difficult to estimate intake of individual food items when foods are grouped Provides little information on food preparation methods or specific details about foods Tends to overestimate energy and some nutrients (extremely high nutrient estimates are not uncommon) Nutrient estimates often require adjustment for caloric intake

OCR for page 57
Dietary Risk Assessment in the WIC Program Criterion or Characteristic Diet History 24-Hour Diet Recall Validity (accuracy) Validity is difficult to assess since intakes cannot be independently observed; recall period may be difficult for subject to conceptualize; less error from withinperson variability, but error of methods has not been quantified Standardized methodology available; provides valid estimates of mean nutrient intakes for groups, but not for individuals Well-defined time period; accuracy depends on subject’s or proxy’s recall or memory Does not alter person’s dietary habits Portion sizes may be difficult to estimate accurately; may be more difficult to assess young children’s diets since more than one proxy respondent may be required to report the day’s intake completely Potential for systematic bias Reliability (reproducibility) Recall of past diet may be influenced by current diet Higher energy intakes in first vs. subsequent administrations in children ages 5–18 yr Repeated diet history shown to be reproducible based on 1-mo diet history and 24-h urinary nitrogen excretion Day-to-day variability in an individual’s intake reduces reliability of a single day’s or few days’ intake

OCR for page 57
Dietary Risk Assessment in the WIC Program Food Record Food Frequency Questionnaire (FFQ) Captures more than one day of intake Portions can be weighed or measured for improved accuracy Validity can be improved by instruction and monitoring Completeness of recording decreases as the number of days increases Sequential days are not independent observations; subject may alter intake or not record all food items Potential for systematic bias More useful for qualitative intakes rather than for quantitative intakes Calibrations with diet recalls or food records provide correlations in the range of 0.3–0.6 or 0.7 for most nutrients (mean 0.5), or 0.5–0.8 after statistical adjustment for energy and within-person variation Eating habits are not affected by method Potential for systematic bias Multiple days provide reliable information for less frequently consumed foods Intraclass coefficients range from 0.5 to 0.9 for two 7-d food records Many types of FFQ instruments available Reliability is influenced by heterogeneity of population Less standardized method, especially for infants and young children May require subject to group foods Requires subject to estimate frequencies of intake Correlation coefficients of 0.4–0.7 for food groups and food items Food lists may not contain cultural foods usually eaten Many FFQs have been calibrated (rather than validated) against other methods; some FFQs have been tested against biomarkers Higher energy intakes in first vs. subsequent administrations in children ages 5–18 yr; portion sizes may be unreliable

OCR for page 57
Dietary Risk Assessment in the WIC Program Criterion or Characteristic Diet History 24-Hour Diet Recall Issues relevant to WIC populations Requires a knowledgeable proxy respondent to describe infants’ and preschoolers’ diets Since infants’ and young children’s diets can be variable from day-to-day, it may be difficult for a proxy respondent to accurately estimate intake over a certain period of time Difficult to estimate total intake among breast-fed infants Infants’ and preschoolers’ intakes may require multiple proxy respondents to completely capture all foods eaten at home, day care, preschool, and other places throughout the day Overweight adolescent and adult females tend to underreport total energy intake Standardized methodology facilitates capturing ethnic foods and food preparation methods Respondent burden High respondent burden Takes much more time than other methods to administer Respondents must be highly cooperative Does not require literacy if administered by trained interviewer Low respondent burden Requires less effort on the part of the subject High response rates Resource requirements High Requires highly trained interviewers Medium/high depending upon whether recall is computer-assisted and computer-coded Procedure can be administered by telephone Administration time 1 h or more 20–30 min, on average

OCR for page 57
Dietary Risk Assessment in the WIC Program Food Record Food Frequency Questionnaire (FFQ) May be more difficult to use with low socioeconomic groups, recent immigrants, or young children May require multiple proxy respondents to record all foods eaten at home, day care, preschool, and other places throughout the day, leading to incomplete records Food list may not contain the foods consumed by cultural or ethnic groups and be an incomplete list for the individual Commonly used FFQs have been developed more for the general population and major subgroups, and may not be appropriate for all cultural dietary patterns Extreme reporting by the individual (characterized as very high or very low energy intakes) may render the instrument useless for about 20% of individuals Overestimates energy intake by 50% in children ages 4–6 yr High respondent burden Requires much effort and accuracy by subject Subject must be literate; poorer response rates compared with diet recall and FFQ Low to medium respondent burden, depending on length and whether self-administered High response rates Medium/high Procedure can be automated Requires more editing and processing time compared with diet recall Low Does not require highly trained interviewers May be self-administered May be scored with automated procedures or optically scanned Depends on number of days recorded and subject’s abilities 10–15 min, on average for 60–75 item FFQ

OCR for page 57
Dietary Risk Assessment in the WIC Program quality diet recalls or food records. They estimate usual intake of the individuals in the group by obtaining 24-hour recalls or food records over many days using standardized methods. The results are sometimes called a gold standard against which the accuracy of other methods can be compared, despite the possibility of systematic underreporting as mentioned above. Correlations between estimates from FFQs and two 7-day records are typically in the range of 0.3 to 0.6 for most nutrients (Sempos et al., 1992). After statistical adjustments (deattenuation) for energy intake and within-person variation using data from diet recalls or diet records, correlations reported for FFQs used in research studies range between 0.4 and 0.8 (Block et al., 1990; Blum et al., 1999; Brown et al., 1996; Friis et al., 1997; Robinson et al., 1996; Stein et al., 1992; Suitor et al., 1989; Treiber et al., 1990; Willett et al., 1987). Mean correlation coefficients cluster around 0.5 (Jain and McLaughlin, 2000; Jain et al., 1996; Longenecker et al., 1993). In general, correlations for adolescents between the validation standard and diet method were higher for single diet recalls and diet records than for FFQs (McPherson et al., 2000). In one study among adolescents, correlations between 3-day diet records and serum micronutrients ranged from 0.32 to 0.65 (McPherson et al., 2000). The nutrients being assessed and the number of items on an FFQ can affect the validity of the questionnaire. A 15-item questionnaire designed to determine the adequacy only of calcium intake had a 0.8 correlation with intake determined from a 4-day food record (Angus et al., 1989). Among tools that assessed a broad range of nutrients, the highest correlation coefficients that the committee found for women were those reported by the EPIC Group of Spain (1997) for a 50- to 60-minute diet history interview compared with 24-hour recalls obtained over the previous year. Excluding cholesterol, the correlations ranged from 0.51 for -carotene to 0.83 for alcohol; half were 0.7 or greater. However, even a correlation coefficient of 0.8 reflects a substantial degree of error when examined at the level of the individual (see “Agreement of Results by Quartile and Misclassification,” below). Wei et al. (1999) reported on the use of a modified FFQ to assess nutrients in low-income pregnant women ages 14 to 43 years (see Table 5-2). Fourteen percent of the sample was excluded due to unusually high intakes (above 4,500 calories) indicating probable overestimation problems for a proportion of the population. Unadjusted correlation coefficients ranged from 0.3 for carotene to 0.6 for folate, with a mean correlation coefficient of 0.47, following exclusions. The validity of questionnaires with regard to food or food group intake also is a problem. Little evidence is available concerning the ability of FFQs to estimate intake correctly when servings of foods or food groups (rather than nutrients) are the units of comparison (Thompson et al., 2000). In the study by Bohlscheid-Thomas and colleagues (1997), correlation coefficients between food group intakes obtained from the 24-hour recalls and a subsequent FFQ ranged from 0.14 for legumes to 0.9 for alcoholic beverages. For 9 food groups, corre

OCR for page 57
Dietary Risk Assessment in the WIC Program lations were less than 0.4, for 11 they were between 0.4 and 0.6, and for 4 they were greater than 0.6. Similarly, Feskanich and coworkers (1993) reported a range of 0.17 for “other nuts” to 0.95 for “bananas,” with a mean correlation of 0.6 after adjusting for within-person variation in intake. Field et al. (1998) found correlations of 0.1 to 0.3 for vegetables, fruit juices, and fruits, and 0.4 for fruits and vegetables combined among ninth to twelfth graders between a 27-item FFQ and an average of three diet recalls. In general, these correlation coefficients are not better than those found by investigators studying nutrients rather than foods. Correlations with Usual Intake from Diet Recalls or Food Records—Young Children Few validity studies have been conducted of questionnaires designed to assess the diets of young children (Baranowski et al., 1991; Blum et al., 1999; Goran et al., 1998; McPherson et al., 2000; Persson and Carlgren, 1984). Blum et al. (1999) assessed the validity of the Harvard Service FFQ in Native American and Caucasian children 1 to 5 years of age in the North Dakota WIC Program (see Table 5-2). An 84-item FFQ was self-administered twice by parents, at the first WIC visit and then after the completion of three 24-hour recalls. Correlations ranged from 0.26 for fiber to 0.63 for magnesium and averaged 0.5. Persson and Carlgren (1984) evaluated various dietary assessment techniques in a study of Swedish infants and children. They found that a short FFQ (asked of parents) was a poor screening instrument with systematic biases when used for 4-year-olds. Staple foods such as potatoes, bread, cheese, and fruits were overestimated and sucrose-rich foods such as cakes were underestimated compared with results from food records. Agreement of Results by Quantile and Misclassification A number of researchers question the appropriateness of using the correlation coefficient (Hebert and Miller, 1991; Liu, 1994) or a single type of correlation coefficient (Negri et al., 1994) to assess the validity and reliability of food-based questionnaires because a high correlation does not necessarily mean high agreement. This question is especially relevant to the situation in WIC, where estimation errors are of great concern if they result in the misclassification of individuals with regard to their dietary risk. Another way to examine validity and the potential misclassification problem is to examine results of studies that report agreement of the results by quantile. Robinson et al. (1996) compared results from a 4-day diet record obtained at 16 weeks of gestation with those from a 100-item FFQ obtained at 15 weeks of gestation. They found a range: 30 percent of the women were classified in the same quartile of intake for starch, and 41 percent were in the same quartile for

OCR for page 57
Dietary Risk Assessment in the WIC Program TABLE 5-3 Probabilities of Misclassification of a Reference Ranking in Quintiles, Using an Imperfect Alternative Absolute Difference in Quintile Rank P 0.95 0.9 0.8 0.7 0.6 0 0.674 0.573 0.467 0.403 0.357 1 0.315 0.378 0.403 0.400 0.390 2 0.011 0.047 0.113 0.156 0.184 3 0.000 0.002 0.016 0.037 0.060 4 0.000 0.000 0.001 0.003 0.009   SOURCE: Walker and Blettner (1985). calcium. Eight percent were classified in opposite quartiles for energy, protein, and vitamin E intakes. Friis et al. (1997) found that 71 percent of young women were in the same quintile or within one quintile when comparing intakes from an FFQ and three sets of 4-day food records. On average, 3.8 percent were grossly misclassified into the highest and lowest quintiles by the two methods. Freeman, Sullivan, et al. (1994) compared a 4-week FFQ (either the Block FFQ or the Harvard Service FFQ) with three 24-hour diet recalls conducted by telephone among 94 children and 235 women participating in WIC (see Table 5-2). Most correlations between the FFQ and the average of three recalls were below 0.5. The FFQ performed more poorly among children than among women and also among Hispanics than among African Americans and non-Hispanic whites. Suitor et al. (1989) compared the results of three 24-hour dietary recalls and a 90-item FFQ among pregnant women and found that fewer than half of the women who were in the lowest quintile by one method also were in the lowest quintile by the other method (see Table 5-2). The quintile agreement ranged from 27 percent for iron to 54 percent for calcium. Percentage agreement improved (to 43 percent for protein and to 77 percent for calcium) when individuals from the first and second quintile of the FFQ were compared with those in the first quintile of the 24-hour dietary recalls. Clearly, substantial misclassification of nutrient intake occurred at the individual level. Different questionnaires give different results with the same subjects (McCann et al., 1999; Wirfalt et al., 1998). Although McCann and colleagues (1999) reported that the results of different methods are correlated (i.e., r ranges from 0.29 to 0.80), the methods would likely classify individuals differently. Walker and Blettner (1985) examined potential agreement when results from an imperfect method of dietary assessment (e.g., an FFQ) are compared with those from a method believed to be accurate (e.g., many days of research-quality food records). Table 5-3 shows their calculations of the probabilities of misclassification in quintile ranking for correlation coefficients ranging from 0.0 to 0.95.

OCR for page 57
Dietary Risk Assessment in the WIC Program   0.5 0.4 0.3 0.2 0.1 0.0 0.321 0.290 0.263 0.240 0.219 0.200 0.379 0.367 0.355 0.344 0.332 0.320 0.203 0.216 0.225 0.232 0.234 0.240 0.081 0.101 0.118 0.134 0.148 0.160 0.017 0.027 0.038 0.051 0.065 0.080 Note that even if the correlation coefficient between the two methods were 0.8 (ordinarily considered to be excellent correspondence), less than half of all respondents would be allocated to the same quintile by the two methods. This indicates that FFQs hold great potential for misclassification at the level of the individual—regardless of whether nutrient, food, or food group intakes are being estimated. Another way to examine the error in misclassification would be to consider the sensitivity and specificity of the tool and how they would translate to numbers of people miscategorized. Using the relatively high sensitivity and specificity values from the example in the following section and assuming that 25 percent of the population meets the Dietary Guidelines (a value much higher than currently estimated), we see in Table 5-4 that roughly one-fourth of the population (275/1,000 individuals) would be misclassified. Increasing the sensitivity by increasing the cut-off would increase the number of eligible individuals who test positive and reduce misclassification. If a lower, more realistic value representing the percentage of the population that meets the Dietary Guidelines were used, the percent of eligible persons who would be found ineligible would be larger (Table 5-5). Limitations and Uses of Brief Dietary Methods Shortening and simplifying FFQs may make it easier for WIC clientele to respond (whether the FFQ is self-administered or administered by WIC personnel) (Subar et al., 1995), but is the validity of short FFQs acceptable? Based on studies by Byers et al. (1985), Caan et al. (1995), Haile et al. (1986), and others, it is unreasonable to expect that a shortened FFQ will be more accurate than a longer version. For example, Caan et al. (1995) evaluated the sensitivity, specificity, and positive predictive value of a 15-item fat screener when used to identify persons with total fat intakes greater than 38 percent of calories. When they compared results with those obtained from the 60-item Health Habits and

OCR for page 57
Dietary Risk Assessment in the WIC Program TABLE 5-4 Results from a Dietary Tool with a Relatively High Sensitivity and Specificity when 25 percent of the Population Meets the Dietary Guidelines   Result from Dietary Tool Eligible Ineligible Total Does not meet the Dietary Guidelines 563 187 750 Meets the Dietary Guidelines 88 162 250 Total 651 349 1,000 NOTE: Assumptions: sensitivity = 75%, specificity = 65%. TABLE 5-5 Results from a Dietary Tool with a Relatively High Sensitivity and Specificity when 5 percent of the Population Meets the Dietary Guidelines   Result from Dietary Tool Eligible Ineligible Total Does not meet the Dietary Guidelines 713 237 950 Meets the Dietary Guidelines 17 33 50 Total 730 270 1,000 NOTE: Assumptions: sensitivity = 75%, specificity = 65%. History Questionnaire (Block et al., 1990), the fat screener had a low rate (2.7 percent) of gross misclassification—for example, the rate when the lowest quintile by the FFQ was compared with the highest two quintiles by the screener. Caan and colleagues (1995) found that the fat screener had insufficient sensitivity and specificity to be used as a single assessment method for fat. For example, when sensitivity was 75 percent, specificity was 65 percent; but when the cut-off point was raised, sensitivity was 47 percent and specificity was 89 percent. They suggested that the screener would be useful in combination with other dietary methods that also estimate energy intake. Others have found that measures taken to shorten and simplify questionnaires reduce their validity in the research setting. For example, Schaffer and colleagues (1997) reported that median energy intake from a shortened telephone version of an FFQ was 23 percent lower in women than that obtained from a longer FFQ. These investigators reported correlation coefficients ranging from 0.45 for vitamin E to 0.78 for fiber for the two FFQs, suggesting considerable lack of agreement. Similarly, Thompson and coworkers (2000) reported that both a 7-item and a 16-item screener for fruit and vegetable consumption underestimate intake.

OCR for page 57
Dietary Risk Assessment in the WIC Program Brief dietary tools have varying degrees of usefulness, depending upon the need for quantitative, qualitative, or behavioral data. They have been developed to measure usual intake, to screen for high intakes of certain nutrients (e.g., total fat, iron, calcium), or to measure usual intake of particular food groups (such as fruits and vegetables). Several examples have been published (e.g., Block et al., 1989; Caan et al., 1995; Feskanich et al., 1993; Kristal et al., 1990; McPherson et al., 2000; NCHS, 1994; Thompson and Byers, 1994). A major limitation of using them to assess intake in the WIC clinic is that they usually target one nutrient or food group, rather than the entire diet. Thus, they are not directly relevant to determining whether the individual met the Dietary Guidelines or consumed an adequate diet, but they may be useful for planning targeted nutrition education. METHODS TO COMPARE FOOD INTAKES WITH THE DIETARY GUIDELINES The committee was given the charge of investigating methods to determine if an individual fails to meet the Dietary Guidelines. For example, can a practical, accurate method be found or developed to compare reported food intake with recommendations derived from the Dietary Guidelines (USDA/HHS, 2000). The committee found no studies that directly examine the performance of dietary intake tools used to compare an individual’s food intake with the Dietary Guidelines, but did find the following related information. Dietary Intake Form Method Strohmeyer and colleagues (1984) claimed that a rapid dietary screening device (called the Dietary Intake Form, or DIF) “ … provides a rapid, valid, reliable, and acceptable method of identifying the individual with a poor diet” (p. 428). Although the DIF was developed before the existence of the Dietary Guidelines and the Food Guide Pyramid, it was intended to compare a person’s intake with reference values that are similar to the Pyramid’s five food group recommendations. The DIF asks the person to write the number of times the following foods are consumed per week: yogurt and milk; cheese; fish, eggs, meat; dried peas and beans; leafy green vegetables; citrus fruit; other fruits and vegetables; bread; and noodles, grains, and cereals. It also asks the respondent to circle his or her portion size as it compares with specified standards. The average time to complete the DIF is about 4.5 minutes, with a range of 2 to 10 minutes.5 A staff member computes a DIF score by a series of arithmetic processes. The methods that Strohmeyer and colleagues used to test reliability and validity are of questionable relevance to the WIC setting. They tested reliability 5   It is notable that 21 percent of the subjects did not complete the forms; reasons were not reported.

OCR for page 57
Dietary Risk Assessment in the WIC Program using 40 college students who completed the DIF on two occasions 2 weeks apart. Correlation coefficients for the paired food-group and total dietary scores averaged 0.81. Validity testing involved input by researchers rather than by clients and scoring by researchers rather than by clinic staff. Researchers entered data from 29 8-day food diaries onto DIFs and then computed dietary scores. Subsequently, they correlated those DIF scores with total mean Nutrient Adequacy Ratios (NAR, in which NAR equals the subject’s daily intake of a nutrient divided by the Recommended Dietary Allowance of that nutrient). Under these carefully controlled conditions, the correlation of DIF and NAR scores was 0.83. It is likely that reliability and validity testing using the clinic population and clinic personnel would produce less favorable results. More importantly, the limitations described for brief FFQs would be applicable to the DIF as well. Mean Adequacy Ratio Methods A more recent study examined the sensitivity and specificity of two Pyramid-based methods of scoring nutritional adequacy (Schuette et al., 1996). For both scoring methods, registered dietitians obtained data from 1-day food records. They assigned the reported food items to the five Pyramid food groups and “other” (fats, oils, sugars). In the first method, the score represents the number of food groups from which the person consumed at least the minimum recommended number of servings. In the second method, the score represents the number of food groups from which the person consumed at least one serving. The two types of scores were compared with a mean adequacy ratio (MAR-5)6 based on the subject’s intakes of iron, calcium, magnesium, vitamin A, and vitamin B6 as calculated from the same food record. For the first method, sensitivity was 99 percent but specificity was only 16 percent. That is, the first food group method classified nutritionally inadequate diets as inadequate, but it had extremely low ability to classify nutritionally adequate diets as adequate. For the second method, sensitivity was 89 percent and specificity was 45 percent. Thus, even when the cut-off point was more lenient (as in the second method) the ability to identify the nutritionally adequate diets was no better than chance. Either MAR method would depend on data from one or two 24-hour diet recalls, and thus would be subject to all the limitations of diet recalls presented earlier in this chapter. 6   MAR-5 = average nutrient adequacy ratio (NAR) of the five nutrients. NAR is the nutrient content calculated as a percentage of the RDA and truncated at 100.

OCR for page 57
Dietary Risk Assessment in the WIC Program Estimating the Number of Pyramid Portions Accuracy of Estimation The portion sizes that an individual consumes can make a great difference in the degree to which his or her intake meets the recommendations made in the Food Guide Pyramid. Questionnaires either assume a standard portion size, which may or may not be shown on the questionnaire,7 or the respondent is asked to choose a single average portion size (small, medium, or large). However, two major factors affect the accuracy of portion size estimation: (1) withinperson variability in portion size, and (2) ability to recall portion size (see earlier section, “Portion Size Estimation”). Within-person variability in portion size is greater than between-person variability for most foods and for all the food groups studied by Hunter et al. (1988). That is, for food groups, the range of the variance ratios (within/between) obtained from four 1-week diet records was 1.6 (fruit) to 4.8 (meat) when pizza was excluded (the variance ratio was 22 for pizza). No studies were found that examine the extent to which the portion size used on a questionnaire reflects the individual’s average portion size. Using the U.S. Department of Agriculture Protocol for Portion Sizes Even if portion size has been reported accurately, the consumption of mixed foods complicates the estimation of the number of portions a person consumes from each of the five Pyramid food groups. For example, 1 cup of some kinds of breakfast cereal may be about half grain and half sugar by weight so should be counted as only one-half serving from the breads and cereals food group. To determine the numbers of servings of foods in the five major food groups from diet recalls or records accurately, researchers at the U.S. Department of Agriculture (USDA) developed the Continuing Survey of Food Intake by Individuals (CSFII) Pyramid Servings database (Cleveland et al., 1997). Eighty-nine percent of the foods in this database are multiple-ingredient foods. USDA separated these foods into their ingredients and categorized these ingredients into food groups that were consistent with Pyramid definitions for serving sizes (USDA, 1992). If a woman reported eating chicken pie, for example, the database allows estimation of the servings or fractions of a serving of grains, meat, vegetables, and milk products (if applicable) provided by the specified weight of the pie. This means that the accurate comparison of food group intake with recommended intake would require accurate food intake data collected over a number of independent days together with computerized assignment of food ingredients to food groups. Notably, this method of estimating servings was 7   Often the portion size used is either the median for the population group as obtained from a nationwide survey or a common unit such as one slice of bread.

OCR for page 57
Dietary Risk Assessment in the WIC Program used in two rigorous studies that found that fewer than 1 percent of women (Krebs-Smith et al., 1997) and young children (Munoz et al., 1997) met recommendations for all five food groups (see Chapter 8). Healthy Eating Index Scores USDA’s Center for Nutrition Policy and Promotion developed the Healthy Eating Index (HEI) to assess and monitor the dietary status of Americans (Kennedy et al., 1995). The 10 components of the HEI represent different aspects of a healthful diet. Five of the components cover the five food groups from the Food Guide Pyramid and the other five cover elements of the 1995 Dietary Guidelines concerning fat, saturated fat, cholesterol, sodium, and variety. The computation of the number of servings from each food group requires the use of complex computerized methods to disaggregate mixed foods into ingredients (Cleveland et al., 1997). Each component may receive a maximum score of 10. The index yields a single score (the maximum score is 100) covering diet as a whole and measuring “how well the diets of all Americans conform to the recommendations of the Dietary Guidelines and the Food Guide Pyramid” (Variyam et al., 1998). Theoretically, an HEI score would be a comprehensive indicator of whether a potential WIC participant of at least 2 years of age fails to meet Dietary Guidelines. However, the complexity of methods required to obtain this score limits the feasibility of using it in the WIC setting. The process described above must be used to separate foods into ingredients and categorize the ingredients into food groups, and separate scores must be computed for each of the 10 components of the HEI score. An HEI score of 100 is equivalent to meeting all the Food Guide Pyramid recommendations plus recommendations for fat, saturated fat, cholesterol, and sodium.8 According to Bowman and colleagues (1998), a score of more than 80 implies a “good” diet. Using 1994–1996 CSFII data, approximately 12 percent of the population had a good diet. A good ranking, based on an HEI score of 80, is considerably more lenient than a criterion in which intake of fewer than the recommended number of servings in the Food Guide Pyramid is the cut-off for failure to meet Dietary Guidelines. Even if an HEI score could be obtained accurately in the WIC setting, the score would likely be sensitive, but not specific. The HEI score could be no more accurate than the data from which it is derived. Thus, it is subject to the limitations of the diet recall or FFQ used. 8   The HEI also includes a variety score, but it is not applicable to the current Dietary Guidelines.

OCR for page 57
Dietary Risk Assessment in the WIC Program CONCLUSIONS REGARDING FOOD-BASED DIETARY ASSESSMENT METHODS FOR ELIGIBILITY DETERMINATION Under the best circumstances in a research setting, dietary assessment tools are not accurate for individuals. In particular, a diet recall or food record cannot provide a sufficiently accurate estimate of usual food or nutrient intake to avoid extensive misclassification. Similarly, research-quality FFQs result in substantial misclassification of individuals in a group when results from FFQs are compared with those from sets of diet recalls or food records. Moreover, studies by Bowman et al. (1998), Krebs-Smith et al. (1997), McCullough et al. (2000), and Munoz et al. (1997) (see Chapter 8) suggested that even if the use of research methods were possible in the WIC setting, such methods would identify nearly everyone as failing to meet Dietary Guidelines. In WIC, a dietary assessment method is used by the competent professional authority (CPA) to determine an individual’s eligibility for WIC in the event that the person has no anthropometric, medical, or biochemical risks (see Chapter 2). The result thus may determine whether or not the applicant will receive WIC benefits for a period of several months or longer. Ordinarily, the CPA compares the individual’s reported intake of foods with preset standards for the numbers of servings in five or more food groups. Even if reported intakes were accurate, estimation of food group scores would likely be inaccurate because of the high frequency of mixed foods. If reported intake or assigned food group scores are inaccurate, correct identification of eligibility status is compromised. Shortening FFQs tends to decrease their validity. Very short screens are targeted to one nutrient or food group rather than providing a relatively complete assessment of dietary intake. Methods to compare food intakes with dietary guidance have the limitations of short screens or are too complex to be useful in the WIC setting. Environmental and other factors present in the WIC setting are expected to decrease the validity of tools when compared with those found in the research setting. Consequently, the validity reported for research-quality FFQs can be considered an upper limit for the validity of questionnaires used by WIC. When using these dietary assessment procedures for group assessment, researchers generally have been willing to tolerate a substantial amount of error, for which they could partially compensate by increasing the number of participants in their research or using statistical correction procedures, called corrections for attenuation (Traub, 1994). Error in the assessment of an individual for certification in the WIC program (that is, misclassification error), however, has serious consequences: truly eligible individuals may not be classified as eligible for the services (less than perfect sensitivity), or individuals not truly eligible for the services may receive them (less than perfect specificity). Because of these limitations, the committee concludes that there are not now, nor will there likely ever be, dietary assessment methods that are both suf

OCR for page 57
Dietary Risk Assessment in the WIC Program ficiently valid and practical to distinguish individuals who are ineligible from those eligible for WIC based on the criterion failure to meet Dietary Guidelines or based on cut-off points for nutrient intake. Nonetheless, dietary tools have an important role in WIC in planning or targeting nutrition education for WIC clients, as described in Chapter 9.