Q /‘\/‘\/‘\ || 5 E Texas Agricultural Experiment Station 4,; 5 The lieigfixfirversity System Eftieictwof Socioeconomic i and Demographic Factors on away-from-home ‘l and at-home Consumption of Selected Nutrients Texas Agricultural Experiment Station Edward A. Hiler, Director The Texas A&M University System College Station, Texas \ \'/\'/\‘/ [Blank Page in Orignfl Bulletin] V Texas Agricultural Experiment Station Bulletin 1717 El Effect of Socioeconomic and Demographic Factors on Away-From-Home and At-Home Consumption of Selected Nutrients Rodolfo M. Nayga, ]r., Assistant Professor, Wepartment 0f Agricultural Economics and Marketing, Fnutgers University LIBRARY Oral Capps, In, DEC 0 7 1994 Professor, DflPartment 0f Agricultural Economics, TEX/W A&IVI UNIVERSITY Texas AékM University KEYWORDS: nutrient consumption, food away from home, food at home, socioeconomic and demographic factors, A demand analysis. " I ‘he authors wish to thank Karen Kubena, Stephen Fuller, Dan Padberg, and R. Marie Iones for their helpful comments ‘ 5nd suggestions. Contents Page Executive Summary ............................................................................................................................................................... Introduction .............................................................................................................................................................................. .. 3 Literature Review .................................................................................................................................................................... .. 3 Theory and Methods ............................................................................................................................................................... .. 6 Data Source and Description ............................................................................................................................................... .. 11 I Empirical Results ................................................................................................................................................................... .. I13 I Descriptive Analyses of Nutrient Intakes ................................................................................................................... .. 15 A Regression Analyses of Nutrient Intakes .................................................................................................................... .. 18 Conclusions ............................................................................................................................................................................ .. 22 Literature Cited ...................................................................................................................................................................... .. 23 Appendix A. Means of Average Daily Nutrient Intake for Individuals Across Population Groups ....................... .. 25 Appendix B. Means of Nutrient Intakes as a Percentage of the RDA ........................................................................... .. 3’ Appendix C. Parameter Estimates of the Nutrient Demand Models ........................................................................... Executive Summary The objective of this research is to determine the factors affecting individual intake of nutrients derived from either all food, food away from home (FAFH), or food at home (FAH). Such an evaluation provides a means of assessing the nutritional or dietary quality of American diets as well as provides a comprehen- sive description and understanding of nutrient consumption patterns not only from total food consumption but also from FAFH and FAH. The data used in this study is the Individual Intake phase of the 1987-88 National Food Consumption Survey of the United States Department of Agriculture. The information derived from this study is useful for nutrition educators and policy makers in focusing their nutrition education programs and dollars in both the FAFH and FAH markets. The results of this study generally indicate that average nutrient intakes from FAFH are lower than average nutrient intakes from FAH except for fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids. Interestingly, the results also indicate that roughly 30 percent of the food energy kilocalories comes from FAFH, while the remaining 70 percent comes from FAH. These results have important implications for the away-fnom-home food indus- try. The fast food industry, for instance, has been criticized for serving foods that are "unhealthy" because of high fat content. An example would be a fast food meal of a hamburger, french fries, and milkshake, which contains approximately half the recommended dietary allowance (RDA) of calories and protein for the adult male but supplies only about one-third of the RDA of vitamin C, thiamin, and niacin, and even lesser amounts of iron, calcium, vitamin A, and riboflavin (Putnam and Van Dress, 1984). Concern about health and fitness has encouraged consumers to prefer foods perceived as ”fresh" and "light." In a 1988 Gallup poll conducted for the Na- tional Restaurant Association (1988), almost 60 percent of adult consumers claimed to be very interested in nutrition-conscious menu items in the away- from-home industry. As more consumers demand menu items with improved nutritive value, FAFH operators, to stay competitive, will have to adapt to these nutrition-conscious patrons. Some of the fast food chains, however, have started to respond positively to the demand of the customers for healthier and low-fat food. An example of this is McDonald's, which recently unveiled their 91 percent fat-free McLean Deluxe burger. Individuals residing in suburban areas have slightly higher intakes of most of the nutrients analyzed from FAFH as a percentage of the RDA than do indi- viduals residing in central cities or nonmetro areas. Likewise, individuals from the South have slightly higher intakes of most nutrients from FAFH as a percent- age of the RDA than do those from other parts of the country. Others having higher intakes from FAFH as a percentage of the RDA are the following: em- ployed individuals compared with unemployed individuals; nonfood stamp recipients compared with food stamp recipients; and those who are not on special diets compared with those who are. Across all sex and age groups, the average intake of nutrients from FAH is generally 50 percent or more of the RDA. ' Except for the intake of calcium and magnesium by certain population groups, the individuals in the sample seem to have generally acquired at least two-thirds of the RDA from all foods. The nutrient consumption regression models have also disclosed some interesting results. For instance, individuals who reside in central cities or suburban areas consume lower amounts of fat and the various fatty acids per 1,000 kilocalories from FAFH than do individuals who reside in nonmetro areas. Significant regional differences are also evident in the consumption of various nutrients from FAFH. For example, individuals from the South seem generally to consume higher amounts of various nutrients except alcohol from FAFH but less from FAH than do individuals from other regions of the United States. Evident in the results is the positive relationship between the weight of an individual and the consumption of various nutrients from FAFH. Employed individuals consume more of various nutrients except alcohol and dietary fiber away from home than do unemployed individuals. In contrast, employed individuals generally consume less of various nutrients at home compared with unemployed individuals. As expected, individuals who are on special diets and individuals who receive food stamps consume significantly less nutrients away from home than do their counterparts. An increase in household size is generally associated with an increase in nutrient consumption from FAH but a decrease in nutrient consumption from FAFH. Moreover, age and income generally significantly affect individual consumption of many nutrients either from FAFH or FAH. In particular, con- sumption of almost all the nutrients from FAFH except for some energy-yielding nutrients decreases (increases) initially with successive increments of age (in- come) and then increases (decreases). The consumption pattern of most nutrients from FAH reverses with initial increases (decreases) followed by decreases (increases) with successive increments of age (income). i. 3 Introduction Recent trends indicate that Americans are eating tut more often and are becoming more interested in Kthe nutritional content of the food eaten away from home. Heightened consumer interest in health and nutrition have increased the need for a complete understanding of either away-from-home or at-home nutrient consumption patterns. Moreover, differences in socioeconomic and demographic factors as well as the unequal distribution of economic and other resources have resulted in variations in nutrient consumption both away from home and at home by individuals. Concerns are also increasing about such nutrients as fats and cholesterol (Borra, 1988). How- ever, the shift in emphasis of nutrition policy toward more concern about overconsumption of certain food constituents does not mean less concern about adequate intake of essential nutrients (Windham et al., 1983). The effect of these changing food consump- tion patterns on food distribution systems and on the nutritional status of the consumer must be examined. A variety of activities and programs attempt to improve, regulate, and change consumption patterns of the population. Some of these programs/ activities we direct intervention programs (food stamps, school \. a . nch, school breakfast), food advertising, food labeling, nutrition education and research programs, nutrition surveys, and health care programs. Numer- ous studies have assessed the impacts of these various programs/ activities on the dietary and nutritional status of either an individual or household (e.g., Capps and Schmitz [1991] for a literature review of these studies). - Food consumption relationships are traditionally specified between socioeconomic factors and quantity or expenditure measures. These approaches, however, do not allow inferences regarding the nutritional status of an individual’s diet. Large quantities of food consumed or large expenditures on food either away from home or at home may not necessarily mean adequate nutrient consumption (Adrian and Daniel, 1976). To quote McCracken and Brandt (1987), "... knowledge is the key to rapid and efficient adjust- ments in the food system to changing consumer demands. Along with the impact of these changing food consumption patterns on the food distribution system itself, the nutritional intake of consumers is likely also to be affected." Knowledge of factors that fiffect nutrient demands can also be used in the design ind implementation of nutrition programs. Previous studies on nutrient consumption analy- sis provided baseline information for a number of critical purposes, including the assessment of dietary status and trends and the development of nutrition education policies and public nutrition programs. However, although considerable literature exists on demand models for nutrients, little attention is paid to the analysis of the demand for nutrients derived from either FAFH or FAH. Evaluation of the nutrients that are consumed from either FAFH or FAH would provide a means of assessing the nutritional quality of American diets. Such an evaluation would also allow a comprehensive description and understanding of nutrient consumption patterns not only from total food consumed but also from FAFH and FAH. Our research attempts to fill these voids by using the Individual Intake phase of the 1987-88 National Food Consumption Survey (NFCS). Information on the nutrient intake of individuals in the away-from- home and at-home markets is available from this particular data set. Our research attempts to assess the impact of various socioeconomic characteristics of the individual not only on total nutrient consumption but also on nutrient consumption away from home and at home. Continual changes in the nature and complex- ity of the U.S. food supply also necessitate periodic monitoring of consumption levels of various nutrients either from all foods, FAFH, or FAH. This specifica- tion would allow not only the assessment and com- parison of demand for nutrients derived from FAFH and nutrients derived from FAH but also a compari- son of the results with previous studies. Knowledge of the various socioeconomic and demographic influences on the demands for the nutrients derived away from home can be used as an aid by the FAFH industry to develop marketing programs and strategies. Literature Review This section discusses some studies in the area of nutrient demand, focusing particularly on the differ- ent sociodemographic and economic characteristics. A summary of the selected studies of the demand for nutrients is also presented at the end of this section. Cross-sectional data sets are used in the various selected studies. Some of the existing models are typically used to predict the nutrient intakes of individuals, given economic and sociodemographic characteristics. These studies are also used to assess dietary quality and the impacts of various govern- ment programs on the levels of nutrient intakes (e.g., Lane, 1978; Price et al., 1978; Davis and Neenan, 1979; Akin, Giulkey, and Popkin, 1983; Chavas and Keplinger, 1983; Scearce and Jensen, 1979; Devaney and Fraker, 1991; Basiotis et al., 1983). Common nutrients considered are iron, vitamin A, vitamin C, and calcium. The sociodemographic factors com- monly included in the analyses are income, house- hold size, ethnicity, and urbanization. The influence of these sociodemographic factors on nutrient levels varies across samples and model specifications. Evidence also exists, in most instances, to indicate that participation in government food assistance programs (e.g., Food Stamp Program; National School Lunch Program; National School Breakfast Program; Women, Infants, and Children Program) leads to increases in the levels of nutrient intakes, all other factors held constant. The objectives and general results of the studies mentioned above are subse- quently described in more detail in the following text. In 1976, Adrian and Daniel used the 1965-66 NFCS data set to estimate the impacts of socioeco- nomic characteristics of the households on the con- sumption of protein, carbohydrate, fat, vitamin A, calcium, iron, thiamin, and vitamin C. The socioeco- nomic factors considered were income, degree of urbanization, race, educational attainment of the homemaker, stage of the household in the family life cycle, family size, meal adjustment, and employment status of the homemaker. Regression analysis yielded indications that income positively affects the con- sumption of all the nutrients analyzed except carbo- hydrate. Nutrient consumption responsiveness to income, however, was small, which implies that future income increases have small impacts on food consumption, at least in the United States. The authors also found that urban and rural nonfarm households consume smaller quantities of all nutri- ents analyzed except vitamins A and C than do farm households. Black households also consume less carbohydrate, thiamin, and calcium than do either white or other race households. Families with more educated housewives tend to serve meals with less carbohydrate and fat and more vitamin C. However, households with employed housewives consume more carbohydrates and fats. The authors concluded that food and nutrition programs could be more effective if directed toward urban, less educated, and black households. Lane (1978) examined the effects of the Food Distribution Program and Food Stamp Program on food consumption and nutrient demands in rural California. She used data from a survey of numerous households and considered food energy and eight nutrients (protein, calcium, vitamins A and C, iron, niacin, riboflavin, and thiamin) to be essential for \J adequate nutrition. Tobit analysis was used to deter- mine the relationship between nutritional achieve- ment ratios and the dollar value of food purchased and received per person for programparticipants and nonparticipants. Results indicate that additional values of food available affect nutritional adequacy for both participant and nonparticipant households of the Food Distribution Program. The value of food available was also associated with significantly higher ratios for calories and calcium under the Food Distri- bution Program and significantly higher ratios for calcium, vitamin A, and riboflavin under the Food Stamp Program. Generally, participants in the Food Distribution Program received more food than they would have purchased with food stamps and had slightly higher achievement ratios for nutrients like protein, iron, and thiamin than did nonparticipants. Using a data set for Washington State schoolchil- dren, Price et al. (1978) developed a linear model of demand for 10 nutrients and determined whether either income and the National School Lunch Pro- gram, the National School Breakfast Program, or the y Food Stamp Program affect demand. The nutrients \J analyzed are thiamin, riboflavin, niacin, protein, calcium, phosphorus, iron, vitamin A, vitamin C, and food energy. Their regression models using ordinary least squares reveal that full participants in the School Lunch Program have significantly higher intakes for 5 of 10 nutrients than non- or partial participants. On the other hand, food stamp recipients did not have significantly higher nutrient intakes than did other schoolchildren. The authors also found that food expenditures, income, and the Food Stamp Program did not affect demand as did household size and the National School Lunch Program. Another study, which likewise considered the impact of Food Stamp and nutrition education programs on food group expenditure and nutrient intake of low-income households, was conducted by Davis and Neenan in 1979. The 1976 Expanded Food and Nutrition Education Program (EFNEP) records and a cross-sectional survey of low-income house- holds in Florida data are used in the analyses. Using ordinary least squares, the authors provided estimates of the nutrient adequacy ratios (NAR) for five nutri- ents: protein, calcium, iron, vitamin A, and vitamin C. NAR is equivalent to the ratio of the dietary intake of. nutrient and the RDA for nutrient. The sociodemo- graphic variables used in these NAR equations are the following: income, household size, life cycle family mmposition, ethnicity, employment, urbanization, "-~j*._ 1d education. A similar study was conducted by Scearce and Jensen in 1979 to determine the effects of the Food Stamp Program on the amount of food energy and eight nutrients (protein, calcium, iron, vitamin A, thiamin, riboflavin, niacin, and vitamin C) purchased by low-income families in the southern United States. Scearce and Jensen used the 1972-73 Consumer Expenditure Survey of the Bureau of Labor Statistics (BLS) and a double logarithmic functional form; the results indicate that the food stamp participant families purchase a greater amount of food energy, protein, calcium, iron, vitamin A, and vitamin B1 than do low-income families with similar socioeconomic characteristics that do not participate in the program. These results are consistent with the results reported by Lane (1978) for a California county. Scearce and Jensen (1979) also found that nutrient demands are affected by income, household size, education, life cycle stage, and Food Stamp Program participation. Nutrient demands, however, are not ~» e findings of Chavas and Keplinger (1983). Using fiaftfected by location or race. These results differ from \* ' t e 1977-78 NFCS data set, Chavas and Keplinger examined the effects of nutrition programs on nutri- ent intake of persons from low-income households in the United States. Attention was given to the interac- tion between the effects of these programs and economic and sociodemographic variables. Among the important results were the following: (1) indi- vidual nutrient intake decreases significantly with size of household; (2) nutrient intake is responsive to income for average-income families but not for low- income families; (3) nutrient intake is positively related to the education of the household head; and (4) domestic programs like the WIC program, meal service for the elderly, Food Stamp Program, and the National School Breakfast Program exhibit signifi- cantly positive influences on intake of nutrients. The School Lunch Program is the least effective of the nutrition programs analyzed. Akin et al. (1983) used the 1977-78 NFCS data set to examine the effect of the National School Lunch Program participation on the intake status of partici- pants for four nutrients (vitamin B6, iron, vitamin A, and vitamin C) and food energy. Using switching high-income children. This impact is, however, greater for low-income children than high-income children. Age, economic need, and etlmicity but not location had significant effects on nutrient consumption. Windham et al. (1983) explored the relationships between some demographic and socioeconomic char- acteristics of individuals and the nutrient density of their diets. Using the 1977-78 NFCS data set, regres- sion analysis was employed to test the effects of vari- ous socioeconomic and demographic factors on the nutrient density consumption of food energy and 14 nutrients. The nutrients analyzed are protein, fat, carbohydrate, calcium, iron, magnesium, phosphorus, vitamin A, thiamin, riboflavin, niacin, vitamin B‘, vitamin Bu, and vitamin C. Results from the multiple regression analysis indicate that income had no statis- tically significant effect upon the nutritional quality of diets. However, results of this study also indicate that several socioeconomic characteristics of individuals (e.g., household size, race, geographic region) were related to the dietary density of certain nutrients. Employment status of the male and female head of the household was not significantly related to indi- vidual nutrient density consumption of 12 or 13 nutrients analyzed. Furthermore, the socioeconomic factors included in the models had no significant effects on the consumption of protein, iron, phospho- rus, riboflavin, and vitamin B12 per 1,000 kilocalories. Nutrient density consumption patterns of different socioeconomic groups appeared to be generally uniform. Using data from the spedal low-income component of the 1977-78 NFCS data set, Basiotis et al. (1983) investigated nutrient consumption patterns of low- income households. In particular, the authors exam- ined the relationships between household nutrient availability, food costs, food stamp participation, and selected socioeconomic factors in a simultaneous equations context. The nine-equation model consists of an equation for food cost and eight equations for nutrient availability levels. Aside from food energy, the nutrients examined are protein, calcium, iron, riboflavin, thiamin, vitamin C, and vitamin A. The results indicate that food stamp participation has a positive impact on diet component availability levels. However, the impacts of other socioeconomic vari- ables examined for diet component availability are larger than those for program participation and income. Devaney and Fraker (1989) examined the dietary fiegression analysis, the authors found strong evidence ~_ ‘if a positive impact of the National School Lunch Pro- gram on nutrient consumption by both low-income and impacts of the School Breakfast Program based on 24- hour dietary recall data collected during the 1980-81 school year. Probit analysis was used to estimate a model in which the likelihood of eating breakfast is a nonlinear function of the National School Breakfast Program availability and socioeconomic characteris- tics. Results indicate that the availability of the School Breakfast Program has no effect on the decision to eat breakfast. The authors concluded that the School Breakfast Program is not achieving its basic objective of providing breakfast to children who would not otherwise eat one. Education, employment status of the female head of household, sex, race, age, and ethnicity are significant socioeconomic factors affect- ing the likelihood of eating any breakfast. They also analyzed the effect of the School Breakfast Program on the intake of food energy, cholesterol, vitamin B6, vitamin A, iron, calcium, and magnesium using Heckman's procedure for correcting for selection bias. General findings indicate that participation in the program is associated with higher intakes of calcium but lower intakes of cholesterol and iron at breakfast. Later, in 1991, Devaney and Moffitt used data from the 1979-1980 Survey of Food Consumption in Low-Income Households to assess the dietary effects of the Food Stamp Program. Using ordinary least squares, the authors revealed that the estimated dietary effects of changes in Food Stamp Program benefits are considerably larger than those resulting from changes in cash income. The estimates of the ratios of the marginal propensity to consume for the food stamp benefit and for the cash-income variable range from 3 to 7 across nutrients. These estimates of marginal propensity to consume are comparable in magnitude to those from the Basiotis et al. (1983) study. However, the estimates of the total effects of the Food Stamp Program from the Devaney and Moffitt (1991) study generally are larger than the comparable estimates from the Basiotis et al. (1983) study. Some of the studies that have used the 1977-78 NFCS data set are descriptive in nature. One of these studies was conducted by Windham et al. in 1983. The objective of this study was to evaluate, relative to a standard, the quality of diets consumed by various socioeconomic groups known to have diets with different nutrient content per energy unit of food consumed. The composite data for this study indi- cated that the average American diet based on the 1977-78 NFCS data met or exceeded nutrient density standards for protein, phosphorus, thiarnin, vitamin A, riboflavin, niacin, vitamins B12 and C, and fat per 1,000 kilocalories of food consumed. However, these diets provided only about 80 to 90 percent of nutrient density allowances for calcium, iron, magnesium, vitamin B‘, and carbohydrate. Pao et al. (1985), on the other hand, made com- parisons of the 1-day and 3-day nutrient intakes of y) various individuals in 22 sex-age groups using the 1977-78 NFCS data set. Results indicate that mean energy and nutrient intakes studied from 3-day and 1- day diets yielded similar information on nearly all age groups, regardless of sex. Therefore, 1-day intakes provided nearly the same reliability of results as the 3- day intakes did in the computation of mean intakes of large groups. A summary of selected studies pertaining to the demand for specific nutrients is presented in Table 1. Capps and Schrnitz (1991, p. 22) mentioned that ”knowledge of factors which influence nutrition demands is useful in the design and implementation of programs to promote improvements in nutrition as well as to make current programs more effective and efficient." As in the FAFH expenditure and consump- tion studies, cross-sectional data are used in these selected nutrient demand studies. Common nutrients considered are food energy, iron, vitamin A, vitamin C, and calcium. The sociodemographic factors com- monly included in the analyses are income, house- hold size, ethnicity, and urbanization. . , Some of the existing models are typically used to\J predict the nutrient intakes of individuals. As well, these studies are used to assess dietary quality and the impacts of various government programs on the levels of nutrient intakes. These selected studies, however, have focused their analyses on nutrients derived from total food consumption and have not made comparisons between the nutrients derived from FAH and FAFH consumption. Evaluation of the nutrients that are consumed from either all foods, FAFH, or FAH would provide a means of assessing the nutritional quality of American diets as well as a comprehensive understanding of nutrient consump- tion patterns not only from total food consumed but also from FAFH and FAH. Theory and Methods This section presents the theoretical consider- ations and framework used in developing the nutrient demand models. The basic model of demand for specific nutrients resembles the Engel function, which relates changes in the consumption of a good to changes in income. This function can be derived from consumer theory by assuming that a consumer chooses his consumption bundle to maximize his Table 1. Selected studies pertaining to the demand for specific nutrients. Nutrients Food assistance Sociodemographic flesearcher Data set‘ considered” programs considered‘ factors considered ‘- ; rice, West, Washington 10: Th, Rb, FSP, NSBP, NSLP Household size, Schier, Price State Ni, FE, Pr, Ca, region, urbanization, children Ph, Iron, VA, VC ethnicity Akin, Guilkey, Popkin 1977-78 NFCS 5: FE, VB‘, NSLP Urbanization, (basic sample) Iron, VA, VC income, household Chavas, Keplinger Scearce, Jensen Devaney, Fraker Adrian, Daniel Lane Wavis, Neenan Windham, Wyse, Hansen, Hurst Devaney, Moffitt Basiotis, Brown, Johnson, Morgan school-age children 1977-78 NFCS (spring portion) 1972-73 BLS, CES 1980-81 cross-sectional survey of students, 1980-81 household survey of parents 1965-66 NFCS 1972 survey of households in Kern County, CA (low income) 1976 EFNEP and cross-sectional survey of households in central Florida 1977-78 NFCS 1979-80 Survey of Food Consumption in Low-Income Households (SFC-Ll) 1977-78 NFCS (low income) 12: FE, Pr, Ca, Th, Iron, Rb, VB‘, V8,‘, VC, Ph, VA, Ni FSP, NSLP, NSBP, WlC, Meal service for the elderly 9; FE, Pr, Ca, lrcn, FSP VA, vs, v32, Ni, vc 7: FE, Chol, VB‘, NSBP VA, Iron, Ca, Mg 8: Pr, Carb, Fat, VA, Ca, Iron, Th, VC 9: FE, Pr, Ca, VA, VC, Iron, Ni, Rb, Th FDP, FSP 5: Pr, Ca, iron, VA, VC FSP, EFNEP 15: Pr, FE, Fat, Carb, Ca, lron, Mg, Ph, VA, Th, Rb, Ni, VB‘, V8,‘, VC 11: FE, Pr, VA, VC, Th, FSP Rb, VB‘, Ca, Ph, Mg, Iron 8: lron, Pr, Ca, FE, FSP Rb, Th, VC, VA size, race, ethnicity Income, ethnicity, education, household size, race Education, urbanization, income, life cycle stage, race, household size Race, ethnicity, education, employment status, region, household size, urbanization Income, family size, urbanization, life cycle stage, race, education, employment Income, household size, life cycle stage, family composition, ethnicity, employment, urbanization, education Region, urbanization, income, household size, race, employment, education Income, race, household size, region, ethnicity, urbanization, age Household size 8t composition, urbanization, race, income, region ‘NFCS = National Food Consumption Survey. BLS, CES = Bureau of Labor Statistics, Consumer Expenditure Survey. EFNEP = Expanded Food and Nutrition Education Program. "FE — food energy, Pr — protein, VA — vitamin A, VB, — vitamin B, Th — thiamin, Ca — calcium, VC — vitamin C, VB: — vitamin B‘ Rb — riboflavin, Ph — phosphorus, VB‘ — vitamin B‘, Chol — cholesterol Ni -- niacin, Iron — iron, V8,, —— vitamin Bu, Mg — magnesium Fat — total fats, Carb — carbohydrate °FSP — Food Stamp Program FDP — Food Distribution Program EFNEP — Expanded Food and Nutrition Education Program NSBP — National School Breakfast Program ' NSLP — National School Lunch Program ~ ‘ WlC — Women, Infants, and Children Program Note: A portion of this table was taken from Capps and Schmitz (1991, p. 23). 7 utility subject to a budget constraint. Maximizing a consumer's utility subject to the budget constraint will lead to demand functions for commodities : q = fi(p, y) (1) where p denotes a vector of prices; y is consumer income; and q is the consumption of the ith commod- ity. By extending this model to examine the demand for nutrients, the intake of nutrient k is (Devaney and Fraker, 1989): Nk = Z]. a“ q]. k = 1,...,K (2) where a . denotes the amount of nutrient k (k = 1,2,...,K) contained in each unit of commodity qj. Substituting equation (1) into equation (2) leads to demand functions for nutrients of the following form: . Nk = shiny)- (3) Recognizing that consumers’ preferences may vary with various sociodemographic and anthropomorphic variables and assuming that prices are constant in a cross-sectional data set, a demand model for specific nutrients can be specified as: NH = hgyi, 5) (4) where NH corresponds to the intake of nutrient k by individual i; yi corresponds to the income level of the individual i; and S is a vector representing the various sociodemographic, anthropomorphic, and food programs participation factors. Twenty-eight nutrients are selected for the analy- sis. These nutrients, along with their units of measure- ment, are exhibited in Table 2. From previous studies (see Table 1 for a summary) and conditioned on the data available in the 1987-88 NFCS, the independent variables used in the analyses include urbanization, region, race, sex, employment, household size, age, height, weight, and income. Dummy variables pertaining to whether the individual receives food stamps or not, whether the individual is on a special diet or not, and whether the intake of nutrients occurred mostly during a weekend or a weekday are also included in the analyses. The general model specification used is therefore: N“ = b0 + blurbanl + b2urban2 + baregionl + b,region2 + b5region4 + b6race2 + b,race3 + b8race4 + bghispl + bwsexl + bnemployl + bnfstampl + bndietl + b1 ‘hsize + blsweight + bwheight + bvage + bmagesq + bwweekend + bzoincome + bnincomesq where : NH = average daily intake of nutrient k by individual i. The units of measurement are displayed in Table 2. urbanl = 1 if individual resides in a central city; 0 otherwise urban2 = 1 if individual resides in a suburban area; 0 otherwise regionl = 1 if individual is in the Northeast; 0 otherwise region2 = 1 if individual is in the Midwest; 0 otherwise region4 = 1 if individual is in the West; 0 otherwise race2 = 1 if individual is black; 0 otherwise ‘ race3 = 1 if individual is Asian or Pacific Islander; 0 xj otherwise race4 = 1 if individual is of some other race; 0 otherwise hispl = 1 if individual is Hispanic; 0 otherwise sexl = 1 if individual is male; 0 otherwise employl = 1 if individual is employed; 0 otherwise fstampl = 1 if individual is receiving food stamps; 0 otherwise dietl = 1 if individual is on a special diet; 0 otherwise hsize = household size weight = weight of the individual in pounds height = height of the individual in inches age = age of the individual in years agesq = square of the age of the individual weekend = 1 if the 3-day intake of the individual occurred mostly during a weekend; 0 otherwise income = household income incomesq = square of household income One classification is eliminated from each group of variables for estimation purposes. The base group is individuals who satisfy the following description: reside in a nonmetro area (urban3); in the South ~¢ Table 2. Nutrients used in the analyses and their units of measurement. Nutrient Unit of measurement 1. Food energy kilocalories 2. Protein grams per 1000 kilocalories 3. Total fat grams per 1000 kilocalories 4. Saturated fatty acids grams per 1000 kilocalories 5. Monounsaturated fatty acids grams per 1000 kilocalories 6. Polyunsaturated fatty acids grams per 1000 kilocalories 7. Cholesterol milligrams 8. Carbohydrate grams per 1000 kilocalories 9. Total dietary fiber grams per 1000 kilocalories 10. Alcohol grams per 1000 kilocalories 11. Vitamin A international units 12. Carotene micrograms retinol equivalents (RE) 13. Vitamin E milligrams alpha-tocopherol equivalents (a-TE) 14. Wtamin C milligrams 15. Thiamin milligrams 16. Riboflavin milligrams 17. Niacin milligrams 18. Wtamin B, milligrams 19. Folate micrograms 20. Wtamin B" micrograms 21. Calcium milligrams 22. Phosphorus milligrams 23. Magnesium milligrams 24. Iron milligrams 25. Zinc milligrams I 26. Copper milligrams ‘4. 27. Sodium milligrams 28. Potassium milligrams (region3); white (racel); nonhispanic (hisp2); female (sex2); not employed (employ2); not participating in fie food stamp program (fstamp2); not on a special H 48f (diet2); and the 3-day intake occurred mostly during a weekday (weekday). Household income is used instead of individual income because the NFCS data set provides income information for the house- hold only and not for an individual. The analyses are separated into three different food sources, FAFH, FAH, and all foods eaten, to determine nutrient consumption pattern differences and the factors that affect nutrient consumption across these three food sources. Since excess con- sumption of 1 nutrient does not compensate for deficiencies in another, 28 separate nutrient consump- tion models are specified for each of the 3 food sources to explain nutrient intake. The same set of independent variables is also used for each nutrient consumption model because nutrients are constituent parts of food and therefore may affect the consump- tion of each nutrient analyzed. Each nutrient may be affected differently by the various independent variables included, but there are no a priori reasons to include or exclude any of these factors in any of the nutrient equations. Dietary fiber and the energy- ifi 'elding nutrients, protein, total fat, saturated fatty acids, monounsaturated fatty acids, polyunsaturated fatty acids, carbohydrate, and alcohol are all ex- pressed as nutrient densities or in grams per 1,000 kilocalories to allow proper comparison between individuals. The anthropomorphic measurements of the individual — age, sex, height, and weight -- are included as independent variables to account for physical differences between individuals. For in- stance, male individuals might eat more than female individuals, and taller and heavier individuals might eat more than shorter and lighter individuals. Thus, squared terms are included for income and age to investigate possible nonlinearities in the Engel rela- tionships for FAFH, FAH, and all-food consumption. Hypotheses fonnulated for the nutrient demand part of the analysis are presented in Table 3. Some of these hypotheses are conceived from the results of previous studies. It is hypothesized that individuals who reside in a central city or suburban area demand lesser amounts of most nutrients derived from either all foods or FAH than is demanded by individuals residing in a nonurban area because individuals residing in nonurban areas expend more body energy A _ han do individuals in the urban or suburban areas _ ‘ Adrian and Daniel, 1976). Urbanization is also related to several variables like accessibility to diverse types of stores providing a wide variety of foods; differences in the social, cultural, and economic environment such as occupational opportunities and education; and the amount of information available to the individual (Scearce and Iensen, 1979). Because of the presence in urban areas of numerous restaurants and fast-food facilities, individuals residing in central cities or suburban areas are expected to consume more nutrients from FAFH than do individuals from nonmetro areas. Previous research studies indicate that Southern- ers generally eat more FAFH than do individuals in other regions and therefore might have more intake of nutrients. Race of the individual can affect the pur- chasing habits and hence the amount of nutrients available to an individual. However, race is compli- cated by its relationship with other socioeconomic and demographic characteristics. No a priori hypoth- Table 3. Hypothesized signs of the parameter estimates of most of the nutrient demand equations. Independent variables All food FAFH" FAH" Urbanization (base = nonmetro) Central city - + - Suburban area - + - Region (base = South) Northeast +/- - +1- Midwest +/- - +/- West +/- - +/- Ftace (base = white) Black +/- +/- +/- Asian/Paciiic Islanders +/- +/- +/- Other +/- +/- +/- Ethnicity (base = nonhispanic) Hispanic +/- +/- +/- Sex (base = female) Male + + + Employment (base = unemployed) Employed +/- + - Food stamp participation (base = nonrecipient) v Recipient - - - Special diet (base = not) On special diet - - - Household size +/- - Weight Height Weekend +/- + +/- Age +/- +/- +/- Age squared +/- +/- +/- lncome +/- + +/- lnoome squared +/- - +/- ‘FAFH = food away from home. "FAH = iood at home. Note: The +/- notation denotes no a priori expectations. esis is then specified about the impact of race on the amount of nutrient intake of an individual either from FAFH, FAH, or all foods eaten. Being male is expected to have a greater impact on nutrient intake than being female. Compared with unemployed individuals, employed individuals are hypothesized to consume more nutrients derived from FAFH and all foods but not from FAH. Because food stamp recipients are generally less affluent than food stamp nonrecipients, recipients are expected to consume less amounts of various nutrients than do nonrecipients. Generally, individuals on special diets are expected to consume less amounts of various nutrients than are those not on a special diets. p As household size increases, it is hypothesized that the individual would increase the intake of most nutrients derived from FAH but not from FAFH. The weight and height of an individual are also expected to be positively related to the amount of intake of most nutrients. Individuals who consumed food mostly during a weekend are expected to have greater amounts of intake of most nutrients derived from FAFH than are individuals who consumed food mostly during a weekday. Increases in income are expected to be associated with increases in intake of most of the nutrients from FAFH, but at a decreasing rate. Educational status of the individual is not included in the analyses because the data set provides information only on the educational status of the male and female heads of the households. Because of the nature of the data used (cross- sectional data) in this study, certain parts of the sample have zero FAFH consumption. Depending on the proportion of zero observations on a dependent variable, either ordinary least squares (OLS) or the Heckman Sample Selection Procedure is used in the analysis. However, when the proportion of zero observations on a dependent variable is high, using OLS to estimate the models would result in biased and inconsistent estimates. On the other hand, omitting the zero observations in the OLS runs will result in the estimates characterized by sample selection bias. Heckman (1976, 1979) described sample selection bias as a type of a specification error or omitted-variable problem. Subsequently, Heckman proposed a technique that amounts to estimating the omitted variable using probit analysis and then employing either OLS or generalized least squares to the model with the inclusion of the estimated omitted variable. This technique is appropriate when a notable proportion of the observations on a depen- dent variable is equal to zero. 1O The description of the Heckman Sample Selection Procedure is summarized in Table 4. A two-equation model is specified in equations (1) and (2). The assumptions about the disturbance terms, which a A consistent with random sampling, are specified in equations (3) and (4). Suppose that there are N“ observations for which y“ = 0, and N, (equal to N-No) observations for which y“ > 0. Assuming without any loss of generality that the N, observations for y“ occur first and the sample selection rule that determines the availability of data is as follows: data are available on y“ if y“ > 0 and there are no observations on y“ if y“ 5 0. Then the conditional means of the disturbance terms of (1) and (2) are presented in (5) and (6) of Table 4, respectively. If equation (1) is estimated using the censored or incomplete sample in which only observed data on y“ are included, the omitted vari- able is (ou/ (on? '91,. Consequently, to make the model complete, the specification must be the same as equation (8) in which l“ is inserted as the additional variable to correct for sample selection bias. Table 4. Description of the Heckman Sample Selection Procedure. (1) y“ = x“ B“ + e“ (2) y“=x“|3“+e“ i= 1, N / where (3) E(e“) = 0 forj = 1, 2 (4) E(e“e,.,.) = o“ for i = i‘, j,j' = 1, 2 = 0 for i at i‘ (5) E(e“|y“ > 0) = E(e“|e2, > 42,5“) = (Gm/(Gzzrmlkr E(e“|y“ > 0) = E(e“|e“ > 42,32) = (on/(Gzzrnlxr (6) where (7) l, = f(Z|)/F(Z|) ll = xzBz/(azarn t and F respectively are the density and distribution function of the standard normal distribution. y" = x1531 + ((5,¢)/((522)m))1; + v15 where E(v“) = O (9) E(V"'“) = 0“(1+p'(-Z,7.,-Xf)) where p = o,/(o“ou)‘“ Source: Heckman, J.J. ‘The Common Structure of Statistical Models of Truncation, Sample Selection and Limited Dependent Variabl/ and a Simple Estimator for Such Models,‘ Annals of Economic Social Measurement, 5(1976):475-492. The index zi in equation (7) is obtained from (2) using probit analysis. Equation (2) in this case is *pecified as a qualitative choice model in which a i a .w dependent variable Di is defined such that it takes the value of 1 when yzi > 0 but takes a value of 0 otherwise. By using probit analysis, the probability that ya is positive is determined by the method of maximum likelihood. Consequently, hi can now be computed for each observation of the complete sample using (7) of Table 4 since zi, f(zi), and F(zi) can already be estimated. The variable 7» is called the inverse of Mill's ratio and is defined as the ratio of the value of the standard nomial density function to the value of the standard nomial distribution function. The second step involves the estimation of (8); the estimate of Xi is inserted as a regressor, using either OLS or generalized least squares. Equation (9) of Table 4 implies that the error structure of (8) is het- eroskedastic. A generalized least squares procedure may, therefore, improve the precision of the least square estimates of the complete model (8). The weight that is used for each observation for the gen- eralized least squares procedure is (1 +p’(-zi7ti-).f))‘ / 2. The value of p’ is obtained by regressing each uared residual (vlf) from the OLS estimation of (8) a ailti-kf. Then according to (9), the estimated mtercept of this regression represents the estimated on, while the estimated slope constitutes the estimate of p20". This procedure, developed to correct for heteroskedasticity, however, may break down if the estimated on is negative or the estimate of p’ does not lie in the unit interval. Finally, if the estimated coefficient associated with 7. described above is significantly different from zero, then sample selection bias is said to exist. Otherwise, no sample selection bias arises from using non- randomly selected samples to estimate behavioral relationships (Heckman, 1976). For the nutrient demand equations without a high proportion of zero observations in the dependent variable, heteroskedasticity in the disturbances is checked using the Breusch-Pagan-Godfrey (BPG) test. The BPG test involves an auxiliary regression in which each squared residual from the OLS estimation is regressed on the same set of regressors used in the original equation. An F-test is then performed on all the coefficients except the intercept. The null hypoth- esis of no heteroskedasticity is rejected if the F-test is statistically significant at a specified significance level $0.05 level) in this study. 1r heteroskedastic distur- ances are indeed found in these equations, weighted least squares is employed, the weights being the recip- 11 rocal of the square root of the fitted values from the auxiliary regression involving the squared residuals. Data Source and Description The data set used in this study is the Individual Intake phase of the United States Department of Agriculture's (USDA) 1987-88 Nationwide Food Consumption Survey (NFCS). This data set is the most recent of the national household food consump- tion surveys conducted by USDA. The first USDA food consumption survey was in 1894. Over the years, USDA surveyed food consumption in 1936-37, 1942, 1948, 1955, 1965-66, 1977-78, 1985, 1986, and 1987-88. Data collection for 1987-88 NFCS data set started on April 1987 and continued through August 1988. The 1987-88 survey contains two parts: (1) household food use and (2) individual intake. The household phase provides information on food used by the household for a l-week period and on the cost of that food. The Individual Intake phase, on the other hand, provides 3 days of information on food intake of household members. The Individual Intake phase of the 1987-88 NFCS data set marks only the fifth time that nation- wide information on the dietary intakes of individual household members has been collected by USDA. The 1987-88 NFCS sample was designed using a multistage, stratified area probability sampling method. According to the USDA, the stratification plan took into account geographic location, degree of urbanization, and socioeconomic considerations. This stratification process resulted in 60 strata (17 central city, 28 suburban, and 15 nonmetro areas), which correspond to the geographic distribution, urbaniza- tion, and the density of the population within the conterminous United States. The selection of house- holds for the sample in a particular area was based on a prelisted number of housing units in the area as well as on the estimates of occupancy rates. More details about the data collection process can be obtained from the USDA. The Individual Intake phase of the 1987-88 NFCS data set provides data on 3 days of food and nutrient intake by individuals of all ages surveyed in the 48 contiguous states. These individuals were asked to pro- vide 3 consecutive days of dietary data. The first day's data were collected using 24-hour dietary recall. The period for this 1-day recall was from midnight to 11:59 p.m. on the day preceding the interview (USDA). This collection was done using an in-home personal inter- view. The second and third days’ data were collected using a self-administered 2-day dietary record. Respondents were asked about the sources of each food eaten. Sources included food that was eaten at home, food brought into the home but later eaten away from home, and food that was never brought into the home. USDA considers food from the first two sources to be from the home food supply. Thus, this study considers food from the first two sources to be food at home (FAH) and the third source to be food away from home (FAFH). To get the nutrient intake for the day, USDA summed up the amounts of each nutrient in each food reported by an individual. The amount of each nutrient in each food eaten was calculated using the weight (in grams) of that food and the nutritive value of that food (per 100 grams) from a nutrient data base developed by Human Nutrition Information Service (HNIS). This nutrient data base contains values for food energy as well as 29 nutrients. Aside from food energy, information is available concerning the following nutrients: (1) water, (2) protein, (3) total fat, (4) saturated fatty acids, (5) monounsaturated fatty acids, (6) polyun- saturated fatty acids, (7) cholesterol, (8) carbohydrate, (9) total dietary fiber, (10) alcohol, (11) vitamin A in international units, (12) vitamin A in micrograms retinol equivalents, (13) carotenes, (14) vitamin E, (15) vitamin C, (16) thiamin, (17) riboflavin, (18) niacin, (19) vitamin B6, (20) folate, (21) vitamin Bu, (22) calcium, (23) phosphorus, (24) magnesium, (25) iron, (26) zinc, (27) copper, (28) sodium, and (29) potas- sium. With the exception of water and vitamin A in micrograms of retinol equivalents, all these nutrients and food energy are used in the empirical analyses part of this study. Protein, total fat, saturated fatty acids, monounsaturated fatty acids, polyunsaturated fatty acids, carbohydrate, and alcohol are expressed as nutrients per 1,000 kilocalories or as nutrient densities because these nutrients are sources of energy. The expression of these eight nutrients as nutrient densities accounts for differences in the amount of nutrients taken relative to a set number of calories and therefore allows direct nutrient intake comparison of individuals per 1,000 kilocalories. The individual intake data set also includes information of the individual on the following variables: urbanization, region, race, sex, employment status, food stamp participation, WIC participation, National School Lunch and National School Breakfast Programs participation, special diet information, household size, age, household income, food sources, and foods consumed. 12 The response rate by households in the survey was low. In particular, participation by households drawn into the sample was below 35 percent. This ‘ rate is lower than in previous NFCS data sets. US indicated that a major reason for this occurrence was "heavy respondent burden" in terms of the amount of information asked from each respondent. Survey results may then be biased if respondents and nomespondents have systematically different behav- ior. In terms of the population characteristics in the March 1987 Current Population Survey, the unweight- ed sample represented the population fairly well. Nevertheless, Bethlehem (1988) has shown how re- weighting can reduce the potential for nonresponse bias. Consequently, even if the sample was designed to be self weighting, HNIS and statisticians at Iowa State University created weights for the individuals in the sample to match the characteristics of the sample and the population. Weights were constructed sepa- rately for each of three sex/ age groups and for l-day intakes and 3-day intakes. The three groups that HNIS used were men age 20 and over, women age 20 and over, and persons less than 20 years of age. More information on the weighting procedure is available from HNIS. The number of days in which food intake infor- / mation was available for an individual varied. Thus, for some individuals the information was provided for only a 2-day or 1-day period. The intra-individual effects of variations of the different interview pro- cesses employed in each of the 3 days of intake could be overridden by combining all the individuals with 1 day, 2 days, and 3 days of completed intake. The 1-day dietary recall, for instance, has been criticized because it depends upon memory and may not capture information representing an individual's usual intake. The record method, on the other hand, has been recommended because food consumption can be recorded immediately after ingestion. Respondent burden is, however, increased and intake infonnation may be biased (Pao et al., 1985). Moreover, more than 80 percent of the sample completed 3 days of intake. For these reasons, only individuals who have com- pleted 3-day intakes are included in the analysis. Data are weighted to maintain representativeness of the sample by adjusting for nonresponse. The weights for the 3-day intake are used for this purpose. As in any cross-sectional study, several issues arise in handling the data set. The first data issue is about missing nutrient infomiation on some of the individuals. The dependent variable that is employeqé? T‘ in the nutrient demand part of the study is average daily nutrient intake. Individuals without 3 days of mnpleted intake information and individuals with * a lssing nutrient intake information are then deleted from the sample. The second issue is missing informa- tion on the socioeconomic and demographic vari- ables. This situation constitutes a missing-value problem. The process of obtaining the final samples of observations for the analyses is handled sequentially (Fig. 1). First, the original data, which contains 11,045 individuals, are edited with respect to individuals without 3 days of completed intake. After deleting infomiation from individuals without 3 days of completed intake, the data set contained 8,468 obser- vations. Second, the original data are edited with respect to missing socioeconomic and demographic information as well as missing nutrient intake infor- mation. Three different data sets are then created corresponding to each food source: FAFH, FAH, and all foods eaten. After eliminating data from individu- als having some missing relevant nutrient intake, socioeconomic, and demographic information, the all- foods data set has 6,219 observations. The FAFH and T 186 observations, respectively. flAl-I data sets, on the other hand, have 4,225 and Nutrient Demand Models (Number of Individuals/Observations) 11 ,045 Deleted individuals without 3 days of completed intake 8,468 /l\ Deleted individuals with missing socio- demographic and nutrient intake information / I \ All Food FAFH FAH 6,219 4,225 6,186 Flgure 1. Schematic diagram of the data-screening process. 13 Table 5 presents the descriptive statistics of the 28 nutrients. A listing of the nutrients with their units of measurement is provided in Table 2. On the average, more food energy in terms of kilocalories per day seems to be coming from FAH than from FAFH. In fact, the average amount of food energy an individual gets from FAH per day is about twice the average amount of food energy an individual gets from FAFH per day. No significant differences between FAFH and FAH seem to be evident on the average amounts taken of nutrients that are sources of energy (those expressed as nutrient densities), except for dietary fiber. An individual takes an average of about 3 grams per 1,000 kilocalories of dietary fiber per day from FAH compared with slightly less than 2 grams per 1,000 kilocalories of dietary fiber per day from FAFH. Interestingly, more daily alcohol consumption, on the average, is coming from FAFH consumption than from FAH consumption. In terms of the other nutri- ents, roughly 20 to 40 percent of total nutrient con- sumption comes from FAFH per day. Table 6 shows the means of the exogenous variables used in the regression analyses. Considering the sample used for all foods, about 21 percent of the sample reside in central city areas (urbanl); 49 percent in suburban areas (urban2); and 30 percent in nonmetro areas (urban3). Most of the individuals (35 percent) included in the sample for all foods come from the South (region3). Eighty-six percent are white (racel); 96 percent are nonhispanic (hisp2); 45 percent are male (sexl); 58 percent are employed (employl); 95 percent are nonrecipients of the National Food Stamp Program (fstamp2); 14 percent are on a special diet (dietl); and about 16 percent ate food mostly on a weekend during the 3-day survey period (weekend). Moreover, the average age of the individuals is about 43 years for the all food and FAH samples and roughly 40 years for the FAFH sample. Average household size is approximately three for the three. samples. Average weight is about 159 or 160 pounds and average height is about 66 or 67 inches. Average household income is higher in the FAFH sample than in the all-food and FAH samples. Empirical Results The objective of this research is to determine the factors that affect the demand for various nutrients from all foods, FAFH, and FAH. For this reason, the consumption of 28 nutrients from all foods, FAFH, and FAH are analyzed. This section presents the descriptive results as well as the regression results. Table 5. Descriptive statistics of the nutrients. Mean S. D. 2' Median Max Min. 1. All ma; (N = 6219) U Food energy 1760.53 3694.71 0 1632.21 7131.87 99.94 Protein” 13.94 16.91 0 13.78 35.53 3.15 Total fat” 13.65 13.38 0 13.78 22.24 1.52 Saturated fat” 4.87 6.40 0 4.84 11.28 0.43 Monounsat fat” 5.09 5.61 0 5.13 11.96 0.16 Polyunsat fat” 2.59 4.84 0 2.51 8.35 0.12 Cholesterol 288.89 896.34 3 248.13 1628.07 0 Carbohydrate” 38.40 38.41 0 38.06 83.36 10.61 Dietary fiber” 2.44 5.50 1 2.28 13.23 0 Alcohol” 0.68 10.86 5053 0 22.81 0 Vitamin A(l U) 6018.99 31342.40 0 4054.52 93977.60 14.09 Carotenes 415.49 2713.51 1 231 .90 9328.78 0 Vitamin E 7.82 28.69 0 6.73 71.40 0.14 Vitamin C 89.46 363.52 0 71.49 786.58 0.07 Thiamin 1.32 3.28 0 1.23 8.10 0.06 Riboflavin 1.68 4.39 0 1.53 12.48 0.05 Niacin 19.11 43.34 0 17.71 92.04 0.14 Vitamin B‘ 1.50 3.98 0 1.36 10.50 0.07 Folate 222.17 688.82 0 195.07 1914.36 8.13 Vitamin B" 5.47 45.97 3 3.80 168.34 0 Calcium 703.97 2160.13 0 617.61 3966.81 20.00 Phosphorus 1112.23 2499.90 0 1032.82 4192.67 37.57 Magnesium 233.99 524.38 0 219.53 833.82 20.83 Iron 12.35 33.18 0 11.07 82.92 0.08 Zinc 10.23 32.80 0 9.19 266.06 0.42 Copper 1.05 2.69 o 0.96 14.22 0.02 Sodium 2974.08 7095.98 0 2756.20 12744.30 79.71 Potassium 2393.84 5133.30 0 2274.85 8855.40 197.68 ll. Food away from home (N = 4225) , . Food energy 597.80 2693.85 0 441 .41 3962.32 0.6M Protein” 12.94 30.50 59 12.73 48.45 0 Total fat” 13.87 24.22 143 14.39 31.81 0 Saturated fat” 4.94 10.69 89 4.99 18.59 0 Monounsat fat” 5.17 9.51 143 5.41 13.96 0 Polyunsat fat” 2.68 9.25 90 2.36 16.83 0 Cholesterol 92.10 536.27 238 57.82 858.92 0 Carbohydrate” 38.17 69.04 0 36.49 100.00 0.13 Dietary fiber” 1.91 6.65 183 1.74 25.27 0 Alcohol” 0.93 18.16 3708 0 48.00 0 Vitamin A(IU) 1372.45 12421.70 215 624.37 48601.40 0 Carotenes 101 .27 1022.00 299 34.10 2821 .59 0 Vitamin E 2.45 13.20 153 1.69 21.38 0 Wtamin C 20.47 157.64 281 10.63 261.91 0 Thiamin 0.36 1.81 116 0.25 3.17 0 Riboflavin 0.44 2.19 101 0.31 4.15 0 Niacin 5.82 27.58 71 4.26 38.82 0 Wtamin B, 0.39 1.97 138 0.28 3.06 0 Folate 54.65 283.34 84 37.08 789.22 0 Vitamin Bu, 1.68 27.05 206 0.85 121.57 0 Calcium 198.07 1091.54 6 130.00 2320.02 0 Phosphorus 334.94 1566.86 10 243.34 22511 .95 0 Magnesium 65.09 301 .24 2 48.61 438.84 0 Iron 3.45 17.73 0 2.53 48.29 0.01 ZiflC 3.22 19.82 0 2.22 111.99 0.006 Copper 0.32 1.82 0 0.22 5.94 0.001 Sodium 996.46 4646.79 2 761 .87 7157.26 0 Potassium 696.97 3223.86 20 514.96 4775.12 0 Ill. Food at home (N = 6186) Food energy 1342.96 3546.76 0 1264.27 7131.87 5.60 i Protein” 14.04 20.51 6 13.87 49.62 0 . ' Total lat" 13.26 16.46 1s 19.46 26.09 o ¢ Saturated fat” 4.76 7.67 8 4.72 11 .27 0 14 Table 5. (continued). Mean S.D. Z‘ Median Max. Min. iounsat fat” 4.94 6.93 12 4.99 12.18 0 . olyunsat lat” 2.51 5.53 8 2.42 9.73 0 Cholesterol 224.66 883.26 33 187.24 1554.30 0 Carbohydrate“ 39.30 47.16 0 38.87 1 00.00 4.72 Dietary fiber° 2.58 6.69 24 2.36 13.71 0 Alcohol’ 0.65 14.33 5314 0 45.36 0 Vitamin A(l U) 5072.99 30245.50 21 3254.14 93432 0 Carotenes 345.51 2604.76 36 1 65.75 9303.1 7 0 Vitamin E 6.11 27.91 15 5.04 71.40 0 Vitamin C 75.34 345.51 25 58.73 758.54 0 Thiamin 1.07 3.22 11 1.01 8.10 0 Riboflavin 1 .37 4.36 8 1 .25 12.47 0 Niacin 15.05 43.05 6 14.08 92.04 0 Vitamin B, 1.22 3.98 11 1.11 10.50 0 Folate 184.39 679.45 4 161.04 1914.36 0 Vitamin B" 4.30 41.03 28 2.96 163.27 0 Calcium 566.39 2027.09 2 492.47 3966.81 0 Phosphorus 879.07 2468.32 1 818.99 4192.67 0 Magnesium 188.79 525.46 1 178.57 833.82 0 lron 9.95 33.13 1 8.95 82.92 0 Zinc 7.98 31.11 1 7.30 266.06 0 Copper 0.83 2.78 1 0.76 14.22 0 Sodium 2278.29 6963.17 1 2095.36 12316.60 0 Potassium 1909.19 5159.28 0 1827.44 8759.71 0.44 “Number of observations that are zeros. “Expressed as nutrient densities or in grams per 1,000 kilocalories. Note: These statistics are weighted to make the sample representative of the population. qescriptive Analyses of Nutrient Intakes Appendix A presents the weighted means across socioeconomic and demographic variables of the daily average nutrient intakes of individuals who had completed 3 days of intake. As shown, the nutrient intake means are presented by degree of urbanization, region, race, (Hispanic) origin, sex, employment status, food stamp participation, special diet variable, and week variable. These variables were previously defined in the "T heory and Methods” section. In terms of the amount of food energy intake, the mean daily average for all foods is about 1,700 kilocalories for individuals residing in a central city; about 1,800 for individuals residing in a suburban area; and about 1,750 for individuals residing in a nonmetro area. Roughly 30 percent of the number of food energy kilocalories comes from FAFH and the remaining 70 percent from FAH. Mean intakes of food energy are higher for individuals from the South than for individuals from other regions. Blacks have the highest food energy mean intake from FAFH but the lowest food energy intake from FAH. Hispanics, however, have a lower average daily food energy ' ake from FAFH than do nonhispanics but a higher zrage daily food energy intake from FAH than do nonhispanics. As expected, males have a higher mean of daily average intake of food energy than do 15 females in all three cases (all foods, FAFH, FAH). Employed individuals average about 645 kilocalories per day from FAFH compared with about 470 kilo~ calories per day for unemployed individuals. Em- ployed individuals, however, have slightly lower average food energy intakes per day from FAH than do unemployed individuals. Individuals who are neither receiving food stamps nor on special diets have higher average daily intakes of food energy than do their corresponding counterparts. Individuals residing in the South have lower average intakes of calcium from all foods eaten than do individuals from other regions. Likewise, blacks have a lower average intake of calcium than do whites. These results are consistent with those from the Windham et al. (1983) study, which used the 1977- 78 NFCS data set. Incidentally, these results might be related to previous claims that nonwhites, in general, have low levels of intestinal lactase, which is needed to digest milk sugar lactose to galactose and glucose. In the absence of adequate lactase, lactose remains in the intestine to be fermented by bacterial flora, which results in symptoms of gas, bloating, cramping, and osmotic diarrhea. Average nutrient intakes from FAFH are generally lower than average nutrient intakes from FAH for all demographic variables except total fat per 1,000 Table 6. Means of the Independent variables used ln the nutrient demand analyses. Variable All foods FAFH FAH Urbanization Urban1 0.21 0.20 0.21 Urban2 0.49 0.52 0.49 Urban3' 0.30 0.28 0.30 Region Region1 0.20 0.20 0.20 Region2 0.27 0.28 0.27 Region3' 0.35 0.34 0.35 Region4 0.18 0.18 0.18 Race Racel‘ 0.86 0.89 0.87 Race2 0.10 0.08 0.10 Race3 0.01 0.01 0.01 Race4 0.03 0.02 0.02 Hispanic origin Hispl 0.04 0.0a 0.04 Hisp2' 0.96 0.97 0.96 Sex Sext 0.45 0.47 0.45 Sex2' 0.55 0.53 0.55 Employment status Empioy1 0.58 0.67 0.58 Employ2' 0.42 0.33 0.42 Food stamp participation Fstampl 0.05 0.04 0.05 Fstamp2' 0.95 0.96 0.95 Special diet Diet1 0.14 0.12 0.14 Diet2' 0.86 0.88 0.86 Week Weekend 0.16 0.18 0.16 Weekday‘ 0.84 0.82 0.84 Age 43.33 40.64 43.44 Hsize 3.03 3.02 3.03 Weight 159.48 159.95 159.44 Height 66.72 67.02 66.72 Income 29486.80 32532.10 29472.40 ‘Excluded category in the regression models. kilocalories, saturated fatty acids per 1,000 kilocalo- ries, monounsaturated fatty acids per 1,000 kilocalo- ries, and polyunsaturated fatty acids per 1,000 kilocalories. The intake of fat and the various fatty acids per 1,000 kilocalories is greater from FAFH than FAH for every demographic variable shown in the table except for Asians/ Pacific Islanders and other races and individuals receiving food stamps. Comparison of absolute levels of nutrients with standards for intake is also needed to assess nutri- 16 tional well-being. The weighted means of the nutrient intakes as a percentage of the Recommended Dietary Allowances (RDA) by race and by sex and age gror are exhibited in Tables B.1 and 13.2 in Appendix B. ‘rd RDAs of 16 nutrients included in the NFCS data set are the bases of these percentages. Only 16 of the 28 nutrients used in this report have set RDAs. The RDAs are standards that serve as a goal for good nutrition (National Research Council, 1989). Specifi- cally, the RDAs are recommendations for the average daily amounts of nutrients that population groups should consume over a specific period. The Committee on Dietary Allowances of the Food and Nutrition Board develops the RDA of various nutrients. Ideally the allowance is developed by first determining the average requirement of a healthy and representative segment of each age group for the nutrient under consideration. The variability among the individuals within the group is assessed statistically. The amount by which the average requirement must be increased is then calculated to meet the needs of nearly all healthy individuals. The committee generally sets a recom- mended allowance for nutrients, other than energy, that is above the average scientifically accepted requirement. Consequently, intakes below the recor mended allowance for a nutrient are not necessarilyw inadequate. Some nutritionists consider group averages of two-thirds of the RDA or less to be at risk of malnutrition in some nutrients. However, many researchers, with the release of the 1989 edition of these RDA standards, now consider 100 percent of the RDA values as bases of assessments. Because the RDAs have not been established for all essential nutrients, only the intakes of 16 out of the 28 nutrients as percentages of the RDA are analyzed in this report. As shown in Table B.1 in Appendix B, the average intake of nutrients from FAFH as a percentage of the RDA is smaller than the average intake of nutrients from FAH as a percentage of the RDA. On the aver- age, whites have higher intakes of vitamin A, vitamin E, riboflavin, vitamin Bu, calcium, phosphorus, magnesium, and iron from FAFH as a percentage of the RDA than do other ethnic groups. Blacks, how- ever, have the highest intake of food energy, protein, vitamin C, thiamin, niacin, vitamin B‘, folate, and zinc from FAFH as percentages of RDAs among the ethnic groups on the average. Asians and Pacific Islanders have the lowest average intake of all 16 nutrients, except calcium, from FAFH as a percentage of the RDA but have the highest average intake of vitamirq; and thiamin from FAH as a percentage of the RDA. From food eaten at home, whites have the lowest intake of protein and vitamin C as percentages of RDAs but the highest intakes of vitamin A, riboflavin, Qmin B6, vitamin Bu, and calcium as percentages of As on the average. For whites, the average intake is roughly 20 per- cent or more of the RDA from FAFH for each of the 16 nutrients compared with an average intake of about 60 percent or more of the RDA from FAH for all the nutrients. Averages for blacks are mostly 20 percent or more of RDAs from FAFH. Among Asians and Pacific Islanders, average intakes of food energy, thiamin, vitamin B6, calcium, magnesium, and zinc from FAFH are less than 20 percent of the RDA. Individuals of other races have average intake from FAFH of over 20 percent of RDAs of all 16 nutrients except calcium. From food eaten at home, average intake of almost all nutrients across all the ethnic groups is generally 50 percent or more of the associated RDA. Considering all foods, the average intake of all 16 nutrients for whites is 66 percent or more of the RDA. For blacks, however, the average intakes of calcium and magnesium are below 66 percent of the RDA. Likewise, the average intake of calcium by Asians/ Pacific Islanders is below 66 percent of the RDA. alcium, obtained from milk products, is heavily wmoted by the dairy industry and the government. - wever, a fair proportion of the U.S. adult popula- tion has some degree of lactose intolerance and finds milk products undesirable. Other sources of calcium (fish, sweet potatoes, corn tortillas, some vegetables) are likewise not widely accepted or eaten in large quantities in the United States (Huffman, 1988). These reasons may account for the below 66 percent of RDA average intake of calcium by some of the ethnic groups. These results seem to hold not only among adults but also among children. Price et al. (1978) discovered in their study of Washington State school children that average nutrient intakes of calcium were relatively lower among blacks and Mexican-Ameri- cans compared with those of whites. Table B.2 in Appendix B presents the mean intakes of nutrients as a percentage of the RDA by sex and age groups. These sex and age groups are the same groups used by the Committee on Dietary Allowances of the Food and Nutrition Board in specifying RDAs. As expected, the average intake of all the 16 nutrients from FAFH as a percentage of the RDA is smaller than the average nutrient intakes from FAH as a percentage of the RDA across all sex and tfie groups. On the average, males of ages 19 to 24 .ve the highest average intake of all 16 nutrients from FAFH as a percentage of the RDA except vita- min A, calcium, and phosphorus. Females 51 years of 17 age and over have the highest average intake of vitamin A from FAF H as a percentage of the RDA. Males, on the other hand, 15 to 18 years of age have the highest average intake of calcium from FAFH as a percentage of RDA. Males aged 25 to 50 years have the highest average intake of phosphorus from FAFH as a percentage of the RDA. Across all sex and age groups, the average intake of nutrients from FAH as a percentage of the RDA is generally 50 percent and more. The average intake of nutrients from all foods is also generally 70 percent and more of RDAs by all age groups of males. Average intake of some nutrients from all foods is below 70 percent of RDAs for some of the women groups. For males aged 15 to 18 years, the average intake is 25 percent or more of the RDA for all the 16 nutri- ents from FAFH except vitamin B6 and magnesium. Likewise, the average intake of all nutrients from FAFH is 25 percent or more of the RDA except for vitamin A in males aged 19 to 24 years and for food energy, vitamin B6, and magnesium in males in the 25- to 50-years range. Males who are 51 years or more have average intakes of food energy, vitamin E, vitamin B6, folate, calcium, magnesium, and zinc from FAFH below 25 percent of the RDA. For females, most of the average intake of nutrients from FAFH are 25 percent or more of the RDA except for vitamins E and B6, calcium, magnesium, iron, and zinc in the 15- to 18-years age group; except for vitamins A and B6, calcium, magnesium, and iron in the 19- to 24-years age group; except for food energy, vitamin B6, cal- cium, magnesium, iron, and zinc in the 25- to 50-years age group; and except for food energy, thiamin, vitamin B6, folate, calcium, magnesium, iron, and zinc in the 51 and older age group. Except for vitamin Bu, the average intake of all nutrients from FAFH by pregnant women is below 25 percent of the RDA. Table B.3 in Appendix B shows the weighted means of nutrient intakes as a percentage of RDA for selected population groups. Generally, individuals residing in suburban areas have slightly higher intakes of food energy; protein; vitamins E, C, B6, and Bu; thiamin; niacin; calcium; phosphorus; rnagne- sium; iron; and zinc from FAFH as a percentage of RDA than do individuals residing in either central cities or nonmetro areas. On the average, individuals from either central cities, suburban or nonmetro areas have intakes of 25 percent or more of the RDA from FAFH except food energy, vitamin B6, calcium, magnesium, and zinc. Not many differences of nutrient intakes as a percentage of the RDA, however, are evident between various groups of individuals across the urbanization variables. In terms of regional differences, individuals from the South have the highest intake of food energy, protein, vitamin E, thiamin, riboflavin, niacin, vitamin B‘, folate, phosphorus, magnesium, iron, and zinc from FAFH as a percentage of the RDA compared with individuals from either the Northeast, Midwest, or West. Employed individuals have higher intakes of all the 16 nutrients from FAFH as a percentage of the RDA than do unemployed individuals. In contrast, unemployed individuals have higher intakes of all the 16 nutrients from FAI-I as a percentage of the RDA than do employed individuals. Food stamp nonrecipients have generally higher intakes of all the nutrients from either all foods eaten or FAFH as a percentage of the RDA than do food stamp recipients. Food stamp recipients, however, have slightly higher intakes of food energy, protein, thiamin, niacin, vitamin Bu, and zinc from FAH as a percentage of the RDA than do food stamp nonrecipients. For FAFH, individuals on special diets have lower intakes of all the 16 nutrients as a percentage of the RDA except for vitamin A than do individuals who are not on special diets. On the other hand, individu- als on special diets have higher intakes of vitamins A, E, B6, and C; thiamin; riboflavin; niacin; folate; cal- cium; phosphorus; magnesium; and iron but lower intakes of food energy, protein, vitamin Bu, and zinc from FAH as a percentage of the RDA than do indi- viduals not on special diets. Considering all foods eaten, individuals not on special diets have higher intakes of all 16 nutrients as a percentage of the RDA except vitamins A and C, niacin, folate, and magne- sium than do individuals on special diets. Regression Analyses of Nutrient Intakes Results of the effects of various socioeconomic and demographic variables on intakes of 28 nutrients from 3 different sources — FAFH, FAH, and all foods — are exhibited in Appendix C. Because of the complex influences on individual nutrient consump- tion and the cross-sectional nature of the data, the adjusted R-squares, as expected, are relatively low. The adjusted R-squares across the models range from 0.0172 (vitamin Bu) to 0.2714 (sodium) using all foods data, from 0.0111 (vitamin Bu) to 0.1294 (food energy) using FAFH data, and from 0.0121 (vitamin Bu) to 0.1917 (sodium) using FAH data. The significance level used throughout the analysis is 0.05. A problem in the estimation of regression models using cross-sectional data is heteroskedasticity in the error terms. When heteroskedasticity is present, 18 ordinary least squares (OLS) estimation places more weight on the observations that have large error variances than on those with small error variances r Because of this implicit weighting, OLS parametea estimates are unbiased and consistent, but they are not efficient (Pindyck and Rubinfeld, 1991). To detect the presence of heteroskedasticity, the Breusch-Pagan- Godfrey (BPG) test is used in all the nutrient demand equations except the alcohol regressions where a significant part of the observations in the dependent variable is zero. Details of the BPG test are discussed in the ”Theory and Methods” section of this report. Heteroskedasticity is found in all the 27 regressions analyzed for the all foods data set except vitamin A, carotenes, and zinc equations. Similarly, hetero- skedasticity is found in almost all the 27 regressions using the FAFH data except dietary fiber, vitamin Bu, and zinc equations and in almost all the 27 regres- sions using the FAH data except vitamin A, carotenes, and zinc equations. Weighted least squares (WLS) is used for those equations found to have hetero- skedastic error terms. A substantial part (more than 80 percent of the total observations) of the all-foods, FAFH, and FAH data series has zero intakes of alcohol (see Table 5). In these cases, the use of OLS would result in biased a ,1 inconsistent estimates. Deleting these individuals aw using OLS would not solve the problem of inconsis- tency and would only reduce the efficiency of the estimates because of a smaller sample size. For this reason, the Heckman sample selection procedure is used for each of the three alcohol regressions. Probit analysis is used in the first stage of the Heckman procedure and the dependent variable is given a value of one if there is alcohol intake and a value of zero if there is no alcohol intake. The inverse of Mill's ratio, which would indicate the sample selection bias, is then calculated for each observation and is incorpo- rated as an independent variable in the second stage. In the second stage of the Heckman procedure, OLS is used only on the nonzero observations in the depen- dent variable. No heteroskedasticity tests were performed in the alcohol equations because the error structure of the equation used in the second stage of the Heckman procedure is explicitly heteroskedastic. Consequently, the use of OLS in the second stage of the Heckman procedure produces consistent estimates, but general- ized least squares, when implementation is possible, improves the precision of the estimates. As men- A. tioned, the technique developed to correct for heterw skedasticity in the Heckman procedure may break down when the estimate of on is nonpositive or the estimate of p’ does not lie in the unit interval. Inciden- fly, the estimates of p’ in the alcohol regressions do . lie in the unit interval. Hence, OLS, in this case, is used in the second stage of the Heckman procedure. The coefficient associated with the inverse of Mill's ratio (IMRATIO) is not significant in the FAH but is significant in the all-foods and FAFH models. This significance indicates that sample selection bias could have been introduced in the estimates if the observations with zero alcohol consumption were deleted and not used in the analysis. Adrian and Daniel (1976) noted that urbanization may reflect the impact of various factors like the potential for home food production; accessibility to diverse types of restaurants and stores; distinctions in the social, cultural, and economic environment such as occupational and educational opportunities; and the effects of mass media. Certainly, for FAFH, accessibility to diverse types of restaurants and stores could be a major factor reflected in the urbanization variables. The excluded category in all the regressions is urban3, or individuals residing in nonmetro areas. Significant differences across different degrees of . ' urated fatty acids, monounsaturated fatty acids, . ohol, and vitamin A from FAFH. Individuals who reside in central cities or suburban areas have intakes with lower amounts of fat, saturated fatty acids, and monounsaturated fatty acids per 1,000 kilocalories from FAFH than do those individuals from nonmetro areas. Generally, individuals who reside in nonmetro areas consume more fat, saturated fatty acids, and monounsaturated fatty acids per 1,000 kilocalories from either FAFH, FAH, or all foods than do individu- als who reside in urban areas. The vitamin A regres- sion results, however, tell a different story. Individuals from central cities (suburban areas) consume more (less) amounts of vitamin A than individuals residing in nonmetro areas. These results are consistent with those of Adrian and Daniel (1976). Their results indicate that farm households consume more of all nutrients except vitamin A and C than do urban households. Based on the Heckman procedure estimates, individuals residing in central cities (suburban areas) have higher intakes of alcohol per 1,000 kilocaloriesflfrom FAFH (FAH) than do individu- als residing in nonmetro areas. The joint F-tests conducted indicate that signifi- t regional differences are evident in the consump- in of the following nutrients from FAFH: food energy, fat, saturated fatty acids, monounsaturated firbanization are found in the demand for total fats, 19 fatty acids, polyunsaturated fatty acids, carbohydrate, dietary fiber, vitamin A, carotenes, vitamin E, vitamin C, thiamin, niacin, folate, calcium, phosphorus, iron, sodium, and zinc. The excluded category in all the equations is region3, or the South. Individuals in the Northeast consume more saturated fatty acids per 1,000 kilocalories but less monounsaturated fatty acids per 1,000 kilocalories, polyunsaturated fatty acids per 1,000 kilocalories, dietary fiber per 1,000 kilocalories, vitamin E, and sodium from FAFH than do individuals from the South. Individuals from the Midwest, however, consume more fat per 1,000 kilocalories, saturated fatty acids per 1,000 kilocalo- ries and less food energy, polyunsaturated fatty acids per 1,000 kilocalories, carbohydrate per 1,000 kilocalo- ries, vitamin A, carotenes, vitamin E, thiarnin, folate, and iron from FAFH than do individuals from the South. Individuals from the South also consume more food energy, vitamin E, thiamin, niacin, calcium, phosphorus, iron, zinc, and sodium but less dietary fiber per 1,000 kilocalories and vitamin C than do individuals from the West. Thus, individuals from the South seem to generally consume more amounts of various nutrients from FAFH than do individuals from the Northeast, Midwest, or the West. On the other hand, although the joint F-test for regions in the alcohol regression using FAFH data indicates no significance of the regional variables, the individual effects of the regions are significant. In particular, individuals from the Northeast, Midwest, and the West consume more alcohol per 1,000 kilocalories away from home than do individuals from the South. In contrast, individuals from the South consume less alcohol per 1,000 kilocalories at home than do individuals from other regions. However, individuals from the South generally consume less of the other 27 nutrients from FAH and all foods eaten than do individuals from other regions. Specifically, individu- als from the Northeast and West have higher intakes of saturated fatty acids per 1,0()0 kilocalories, vitamin A, carotenes, vitamin C, and calcium from either FAH or all foods eaten than do individuals from the South. Moreover, individuals from the Midwest have higher intakes of fat and saturated fatty acids per 1,000 kilocalories, calcium, and potassium from either FAH or all foods eaten than do individuals from the South. The results also suggest that some nutrient intakes differ by race of the individual. For instance, races differ significantly in the intake of saturated fatty acids per 1,000 kilocalories, vitamin A, carotenes, vitamin C, niacin, and copper from FAFH. In particu- lar, from FAFH, whites consume less vitamin C and niacin than do blacks; less carotenes than do Asians/ Pacific Islanders; and less vitamin A, niacin, and copper than do individuals of other races. However, whites consume more saturated fatty acids per 1,000 kilocalories, vitamin A, and carotenes from FAFH than do blacks. More significant differences in terms of nutrient intake across the races are evident from the FAH and all-foods models. Incidentally, these differences across races are basically the same for both the FAH models and all-foods models. In general, whites consume more dietary fiber and saturated fatty acids per 1,000 kilocalories, carotenes, vitamin E, thiamin, riboflavin, vitamin B6, calcium, phosphorus, magnesium, and iron than do blacks from either FAH or all foods eaten. Similarly, whites consume more saturated and monounsaturated fatty acids per 1,000 kilocalories and more calcium but less cholesterol, carotenes, and vitamin C from either FAH or all foods eaten than do Asians/ Pacific Islanders. Whites also consume less cholesterol, dietary fiber per 1,000 kilocalories, vitamin C, thiamin, niacin, vitamin B6, folate, and potassium from either FAH or all foods eaten than do other race individuals. These results differ from those of Scearce and Iensen (1979) who found no significant relationship between race and the demand for various nutrients. Hispanics tend to have the same consumption of nutrients from FAFH as the nonhispanics except for copper, of which Hispanics consume significantly less. Hispanics also consume more protein and dietary fiber per 1,000 kilocalories, cholesterol, and folate but less fat, saturated fatty acids, and polyun- saturated fatty acids per 1,000 kilocalories, riboflavin, and calcium from all foods eaten than do non- hispanics. From FAH, Hispanics consume more protein and dietary fiber per 1,000 kilocalories, cholesterol, folate, and magnesium but less saturated and polyunsaturated fatty acids per 1,000 kilocalories and calcium than do nonhispanics. Consumption of alcohol per 1,000 kilocalories; food energy; cholesterol; vitamins E, C, B6, and Bu; thiamin; riboflavin; niacin; folate; calcium; phospho- rus; magnesium; iron; zinc; copper; sodium; and potassium from FAFH is significantly greater for males than females. Males also generally consume more of these same nutrients with the addition of fats and monounsaturated fatty acids per 1,000 kilocalo- ries from either FAH or all foods eaten than do females. Females, however, have higher intakes of fat and all the fatty acids per 1,000 kilocalories from FAF H and more carbohydrate and dietary fiber per 2O 1,000 kilocalories from either FAH or all foods eaten than do males. Employed individuals consume less of only 2 c ~ of the 28 nutrients from FAFH analyzed than do individuals without employment. These nutrients are dietary fiber and alcohol per 1,000 kilocalories. From food eaten at home, however, employed individuals do not consume significantly more of any nutrient than do unemployed individuals. In contrast to the FAFH results, employed individuals consume signifi- cantly less food energy; cholesterol; vitamins A, C, E, and B6; carotenes; thiamin; riboflavin; niacin; folate; calcium; phosphorus; magnesium; iron; zinc; copper; sodium; and potassium from FAH than do unem- ployed individuals. In terms of all the foods eaten, employed individuals consume more fats, monounsaturated, and polyunsaturated fatty acids per 1,000 kilocalories but less carbohydrate and dietary fiber per 1,000 kilocalories; vitamins A, C, and B6; riboflavin; folate; calcium; magnesium; and potassium than do unemployed individuals. Individuals receiving food stamps do not con- sume significantly more of any nutrient from FAFH than do individuals who are not receiving food stamps. Those individuals who receive food stamps,,. however, consume significantly less fat, saturated a l, polyunsaturated fatty acids per 1,000 kilocalories; & food energy; vitamin E; calcium; and phosphorus from FAFH than do those who are not receiving food stamps. In contrast, those who are receiving food stamps consume more alcohol, vitamin Bu, and copper but less carotenes and vitamin C from either FAH or all foods eaten than do food stamp nonrecipients. The results using the all-foods data set are not quite the same as those from the Scearce and Jensen study in 1979. Using the 1972-73 Consumer Expenditure Survey of the Bureau of Labor Statistics, their results indicate that the food-stamp-participat- ing families purchased a greater amount of food energy, protein, calcium, iron, vitamin A, and vitamin B] than did nonparticipating families. Individuals on special diets do not consume significantly more of any nutrient from FAFH except protein, polyunsaturated fatty acids, and dietary fiber per 1,000 kilocalories than do those individuals who are not on a special diet. As expected, individuals who are on a special diet consume significantly less food energy; saturated fatty acids and carbohydrate per 1,000 kilocalories; cholesterol; vitamins E, B6, and Bu; thiamin; riboflavin; niacin; folate; calcium; phos- phorus; magnesium; iron; zinc; copper; sodium; an Y potassium from FAFH than do individuals not on a B U0 10H SIBIIPIAIPUI Op H2111 Hflvfl U101] UHIISSQIOd Nu2 fu1n1pos !10dd00 fau1z £11011 .'111n1s01132111 !sn.1011d 1-s011d !111n10[20 501210; !u1021u !u1A2[;0q11 fu11u21111 !“g pu2 '93 ‘g1 su111121111 .'[0101s0[0110 fs0110[200[p[ 000'[ 10d 0121pA11oq120 pu2 sp102 A112; p0121n12s !A310u0 poo; ss0[ A[1u201;1u81s 0111ns1100101p [2100ds 2 110 012 011M s[2np1A1pu1 'p0100dx0 sv '101p [2100ds 2 uo 1011 012 011M s[2np1111p111 0s0111 op 112111 s0110[200[1>[ 000'[ 10d 10q1; A12101p pu2 'sp102 A112; p0121n12sunA[od 'u101o1d 1d00x0 11,1111; u101; 1110111nu Au2 ;0 010111 A[1u201;1u21s aumsuoa 1011 op s101p [2100ds uo s[2np1111p11[ 001111112; 3u112d101112du0u p1p u2111 ‘g 11111121111 pu2 ‘V u1111211A ‘11011 'u1n10[20 'u10101d 'A310u0 poo; ;01un01112 1012013 2 p0s21101nd s01[11112; 3u1 -12d101112d-du121s-p00; 011112111 01201p111 s1[ns01 110111 ‘s011s11215 10q21 ;0 n201ng1 0111 ;0 A0111n5 011111pu0dxg 181111181100 €L-ZL6I @111 3111811 '6.461 u! K2018 11891181 pu2 0012005 0111 u101; 0s0111 s2 0u12s 0111 011nb10u 012 10s 212p spoo;-[[2 01112111sn s1[ns01 0111 ~s11101d1001110u du121s p00; op u2111 u0120 spoo; [[2 10 Hvd 1011110 11101; 3 11111121111 pu2 s01101o120 ss0[ 1nq 10dd00 pu2 mg u1111211A ‘10110012 010u1 aumsuoa sdu121s poo; 3u1A10001 012 011M 0s0111 '1s2111100 111 -sd11121s poo; 31111110001 10u 012 011M 0s0111 op u2111 115V; 11101; sn1011ds011d p112.'11m10[20fg u111121111 {AS10110 poo; % .'s0110[200[1>[ 000'[ 10d sp102 A112; p0121n12sunA[0d ’ e p0121n12s '12; ss0[ A[111201;1u21s 0111nsu00 ‘10110111011 "sd11121s poo; 01110001 011M s[2np1A1pu10s011_[_ 'sd11121s poo; 3u11110001 10u 012 011M s[2np1A1pu1 op u2111 Hgvg 11101; 1110111n11 Aue ;0 010111 A[111201;111fi1s 0111ns -u001011 op sd11121s poo; 31111110001 s[2np1A1pu[ -s[2np1111pu1 p0A0[du10un op 112111 u1n1ss21od pu2 !111n1s0u82111 f111n10[20 501210; !u11\2[;0q11 5911 PW '0 ‘v 811111121111 1881101200101 000’1 18d my 14121810 pu2 0121pAq0q120 ss0[ 1nq s0110[200[1>[ 000'[ 10d sp102 A112; p0121n12sunA[od pu2 'p0121n12sun0uou1 ‘s12; 010111 0111ns1100 s[2np1A1pu1 p0A0[d1110 'u0120 spoo; 0111 [[2 ;0 su1101 u] 's12np1111pu1 p0A0[d -u10un op 112111 [.1175 u101; 111n1ss210d p112 £111n1p0s !10dd00 £01112 fuo11 .'u1n1s0uS2u1 .'sn101[ds011d !11m10[20 101210; !111021u!1111\2[;oq1.1!u1111211[1 .'s0u010120 59g pu2 "_:[ '3 ‘V su1u1211A .'[0101s0[0110 $310110 poo; ss0[ A[1u20 -1;1u31s aumsuoa s[2np1A1p111 p0A0[du10 's1[ns01 Hgvg 0111 01 1s2111100 111 's[2np1111p111 p0A0[du1011n op u2111 1110111nu Au2 ;0010u1 A[1u201;1u81s aumsuoa 1011 op s[2np11\1pu1 p0A0[du10 '10A0M011 '0u1o1112 u0120 poo; 11101,; 's011o[200[1>[ 000'[ 10d 10110012 pu2 10q1; A12101p 012 s1u0111nu 0s0111 -1u0111A0[du10 11101111111 s[2np1A1pu1 op u2111 p0zA[2u2 1.15115 u1o.1; s1u0111n11 g3 0111 ;0 a z A[u0 ;0 ss0[ 011msu00 s[2np1A1pu1 p0A0[d111§[ 's0[2111 op u2111 u0120 spoo; [[2 10 1.11711 1011110 u101; s0110[200[1>[ 000'[ OZ 10d 10q1; A12101p pu2 0121pA110q120 010u1 pu2 Hgvg 11101; s0110[200[p[ 000'[ 10d sp102 A112; 0111 [[2 pu2 12; ;0 s0>[21u110118111 011211 ‘1011011/1011 's0[2u10:[ 's0[21110; op 112111 110120 spoo; [[2 10 Hvd 1011110 11101; s011 -0[200[1>[ 000'[ 10d sp102 A112; p0121n12sunouo111 pu2 s12; ;0 110111pp2 0111 1111M s1u0111n11 01112s 0s0111 ;0 010u1 011msu00 A[[2.101103 os[2 $012141 's0[2u10; u2111 s0[2111 10; 10120.13 A[111201;11131s s1 H5175 u101; 111111ss210d p112 fum1pos £10dd00 £01112 fu011 !111n1s0u32111.’sn1 -011ds011d !111n10[20 501210; 111102111 !u1A2[;0q11 fu111121111 (“g pu2 '93 '3 ‘g1 su1111211A .’[0101s0[0110!A310u0 poo; .'s0110[200[1>[ 000’[ 10d 10110012 ;0 u011du1nsu03 's01u2ds111u0u op 112111 um10[20 pu2 s0110[200[1>[ 000'[ 10d sp102 A112; p0121n12s11nA[0d pu2 p0121n12s ss0[ 1nq 111111s01132111 pu2 ’012[0; '[0101s0[01[0 's011o[200[1>[ 000'[ 10d 10q1; A12101p pu2 u10101d 010111 011ms1100 s01112ds11_1 ‘Hvd u101_.[ -s01u2ds111 -uou op u2111 u0120 spoo; [[2 u101; 1111110120 pu2 'u1112[;0q1.1 's0110[200[1>[ 000’[ 10d sp102 A112; p0121n12s -unA[0d pu2 'sp102 A112; p0121n12s '12; ss0[1nq 01210; pu2 ’[0101s0[0110 's0110[200[1>[ 000'[ 10d 10q1; A12101p pu2 u10101d 010u1 0111nsu00 os[2 s01u2ds111 'ss0[ A[111201;111B1s 011msu00 s01112ds11_1 11011[M ;o '10dd00 10; 1d00x0 s01u2ds111u011 0111 s2 Hgvd u101; s1u0111nu ;0 u011d111ns1100 01112s 0111 011211 01 pu01s01112ds111 -s1u01.11nu sno112A .10; p112u10p 0111 pu2 0021 u00M10q d11[su0112[01 111201;1u81s ou puno; 011M (6461) 110su0[ pu2 0012005 ;0 0s0111 u101; 10;;1p s1[ns01 0s0111 's[2np1111pu1 0021 101110 op u2111 110120 spoo; [[2 10 [.1105 1011110 11101; 111111ss210d pu2 '012[0; '°g 11111121111 ’u1021u 'u1u121111 '3 u111121111 's0110[200[1>[ 000'[ 10d 10q1; A12101p '[o101s0[0110 ss0[ 0111ns1100 os[2 s01111M -s10pu2[s1 01;102(1/su21sv op u2111 u0120 spoo; [[2 10 1.1V; 1011110 11101; 3 u1111211A pu2 's0u010120 '[0101s0[o110 ss0[ 1nq 1111110120 010111 pu2 s0110[200[1>[ 000'[ 10d sp102 A112; p0121n12sunou0u1 pu2 p0121n12s 010111 011msu00 s01111M '.A[12[_11u15 -u0120 spoo; [[2 10 Hvg 1011110 11101; s>[02[q op u2111 u011 pu2 '111n1s0u32111 'sn1011ds011d ‘1111110120 '93 u111121111 'u11\2[;0q11 'u111121111 ‘g1 111111211A 's0u010120 's0110[200[1>[ 000'[ 10d sp102 A112; p0121n12s pu2 10q1; A12101p 010u1 aumsuoa s01111M '[210u03 u; 's[0p0u1 spoo;-[[2 pu2 s[0p0u1 1.12111 0111 1110q 10; 01112s 0111 A[[e01s2q 0.12 s0021 ss0102 s00u010;;1p 0s0111 'A[[21u0p10u[ 's[0p0111 spoo;-[[2 pu2 Hvg 0111 u101; 1u0p1A0 012 s0021 0111 ss0102 0>[21u1 1u0111nu ;0 s111101 111 s00u010;;1p 1u201;11131s 010w 's>[02[q op u2111 [_[_._1V_:1 11101; s0u010120 pu2 ‘V u1u1211A 's01.10[200[1>[ 000'[ 10d sp102 A112; p0121n12s 010111 01uns1100 s01111M ‘1011011101-1 's0021 101110 ;0 s[2np1A1pu1 op u2111 10ddo0 pu2 '111021u ‘V 11111121111 ss0[ pu2 .’s10pu2[s[ 01;102([ /s1121sV op u2111 s0u010120 ss0[ fs>[02[q op u2111 u1021u special diet. However, from food eaten at home, individuals who are on a special diet consume more fitein and dietary fiber per 1,000 kilocalories; . Johydrate; vitamins A, C, and B6; carotenes; folate; magnesium; and potassium but less fat, saturated and monounsaturated fatty acids per 1,000 kilocalories, food energy, cholesterol, zinc, and sodium than do individuals not on special diets. If all the foods eaten are considered, individuals who are on special diets consume more protein and dietary fiber per 1,0()0 kilocalories, carbohydrate, vitamins A and C, carotenes, folate, and copper but less fat, saturated and monounsaturated fatty acids per 1,000 kilocalo- ries, food energy, cholesterol, thiamin, vitamin Bu, iron, zinc, and sodium than do individuals who are not on special diets. Intakes of individuals during the weekend are significantly higher in vitamin B6, zinc, and copper from FAFH than intakes of individuals during the week. On the other hand, for the FAH, intakes of individuals during the weekend are significantly higher in cholesterol but significantly lower in protein and dietary fiber per 1,000 kilocalories; vitamins A, C, and B6; carotenes; niacin; folate; magnesium; copper; 6 d potassium than intakes of individuals during the fleck. If all the foods eaten are considered, intakes of ..dividuals during the weekend are significantly V greater in food energy, cholesterol, and zinc but significantly less in protein and dietary fiber per 1,000 kilocalories, vitamins A and C, and carotenes than intakes of individuals during the week. A positive relationship is noted between the weight of an individual and food energy, monoun- saturated fatty acids per 1,000 kilocalories, choles- terol, vitamins E and B6, thiamin, riboflavin, niacin, folate, calcium, phosphorus, magnesium, iron, zinc, copper, sodium, and potassium consumption from FAFH. Positive relationships also exist between the weight of an individual and protein per 1,000 kilo- calories, cholesterol, zinc, and sodium consumption from FAH and between weight of an individual and food energy, protein and monounsaturated fatty acids per 1,0()0 kilocalories, cholesterol, thiamin, ribo-flavin, niacin, vitamins B6 and Bu, phosphorus, iron, zinc, and sodium consumption from all food eaten. An inverse relationship, however, is reflected between the weight of an individual and the consumption of the following nutrients: carbohydrate per 1,000 kilo- calories from either FAF H, FAH, or all foods eaten, %etary fiber per 1,000 kilocalories from either FAH or l foods eaten, and polyunsaturated fatty acids per 1,000 kilocalories and copper from all foods eaten. 21 Results generally indicate that an increase in household size is associated with an increase in nutrient consumption from FAH but a decrease in nutrient consumption from FAFH. These results are consistent with the results of Adrian and Daniel (1976) using the 1965-66 NFCS data set. Specifically, a negative relationship exists between household size and protein, polyunsaturated fatty acids, and dietary fiber per 1,000 kilocalories; food energy; cholesterol; vitamins A, E, C, B6, and Bu; carotenes; thiamin; riboflavin; niacin; folate; calcium; phosphorus; magnesium; iron; zinc; copper; sodium; and potas- sium consumption from FAFH. Conversely, a positive relationship is found between household size and fat, saturated and monounsaturated fatty acids per 1,000 kilocalories; food energy, cholesterol, thiamin, ribofla- vin, niacin, calcium, phosphorus, magnesium, iron, zinc, sodium, and potassium consumption from FAH. If all foods eaten are considered, a negative relation- ship is found between household size and polyun- saturated fatty acids, alcohol, and dietary fiber per 1,000 kilocalories; vitamins A, E, C, and B6; carotenes; niacin; folate; calcium; phosphorus; magnesium; copper; and potassium consumption. A positive relationship is found between household size and fat and monounsaturated fatty acids per 1,000 kilocalo- ries and sodium consumption from all foods eaten. Similarly, Chavas and Keplinger (1983) using the 1977-78 NFCS data set found that individual nutrient intake decreases significantly with household size. Age of the individual is generally a significant factor affecting individual consumption of many nutrients either from FAFH, FAH, or all foods eaten. Consumption of almost all the nutrients from FAFH except for some of the energy-yielding nutrients decreases initially with successive increments of age and then increases, as indicated by the significant negative and positive signs of the age and age- squared coefficients, respectively. The relationship of consumption behavior with age of the individual is generally the opposite for foods eaten at home. For instance, as evidenced by the significant positive and negative signs of the age and age-squared coefficients, respectively, consumption of protein, fat, monounsaturated and polyunsaturated fatty acids per 1,000 kilocalories, food energy, cholesterol, carotenes, niacin, magnesium, zinc, copper, and potassium from FAH increases initially, then peaks, and then de- creases with further increments of age of the indi- vidual. Age coefficients from the models for dietary fiber per 1,000 kilocalories, vitamin A, iron, and sodium from FAH are positive and significant, but the age-squared coefficients are not. Less uniformity in the nutrient consumption patterns in relation to the age variable is found using all foods eaten. In particu- lar, a significant positive sign on the age variable and a significant negative sign on the age-squared vari- able is found on protein, dietary fiber, fats, monounsaturated and polyunsaturated fatty acids per 1,000 kilocalories, magnesium, copper,and potassium consumption models. Conversely, a significant negative sign on the age variable and a significant positive sign on the age-squared variable is found on carbohydrate per 1,000 kilocalories, food energy, thiamin, vitamins B6 and Bu, folate, calcium, phospho- rus, iron, and sodium consumption models. Nonlinear consumption patterns are also reflected in some of the significant coefficients of the income and the income-squared variables. Household income is more of a significant factor affecting individual consumption of nutrients from FAFH than from either FAH or all foods eaten as evidenced by the number of significant coefficients. Moreover, the general pattern of nutrient consumption from FAFH typically starts with an increase and then declines with successive positive increments of income, as indicated by the significant positive and negative signs of the income and income-squared coefficients. This consumption pattern is evident in the following consumption models from FAFH: protein, fat, saturated fatty acids, monounsaturated and polyunsaturated fatty acids and dietary fiber per 1,000 kilocalories, food energy, thiamin, riboflavin, niacin, folate, magnesium, iron, and sodium. Consumption models from FAFH with a significant and positive income coefficient but insig- nificant income-squared coefficient are alcohol; vitamins A, E, C, and B6; carotenes; calcium; phospho- rus; copper; and potassium. In contrast, the general pattern of nutrient consumption from FAH starts with a decrease and then increases with successive incre- ments of income. Individual consumption of food energy, cholesterol, riboflavin, niacin, phosphorus, and sodium from FAH depicts this same pattern with increases in household income. In the models for monounsaturated fatty acid per 1,000 kilocalories, phosphorus, iron, zinc, and potassium consumption from FAH, coefficients are negative and significant for the income term but insignificant on the income- squared term. If all the foods eaten are considered, the consumption of carotenes, vitamins E and C, and copper initially increases and then decreases with successive increments in income. The coefficients associated with the income-squared term in the models for polyunsaturated fatty acid and dietary 22 fiber per 1,000 kilocalories, vitamins A and B6, folate, magnesium, and potassium from all foods eaten are not statistically significant although the income ter ~ are positive and significant. In general, these res contrast to those of the Windham et al. (1983) study, which used the 1977-78 NFCS data set. Their results indicate that income had no statistically significant effect upon the nutritional quality of the diets. Conclusions Heightened consumer interest in health and nutrition has increased the need for a complete understanding of nutrient consumption patterns not only from food eaten at home but also from food eaten away from home. Moreover, food consumption relationships are traditionally specified between socioeconomic factors and quantity or expenditure measures. These approaches, however, do not allow inferences regarding the nutritional status of an individual's diet. On the other hand, although consid- erable literature exists on demand models for nutri- ents, little attention is paid to the analysis of the demand for nutrients derived from either FAFH or FAH. In this light, this study evaluates individual nutrient consumption to provide a comprehensive Q description and understanding of nutrient consump- tion patterns not only from total food consumption but also from FAFH and FAH. In addition, the infor- mation obtained from this study could be used as a guide by nutrition educators and policy makers in the design and development of nutrition educational programs. The results in this study generally indicate that average nutrient intakes from FAFH are lower than average nutrient intakes from FAH except for fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids. This information has important implications for the away-from-home food industry. The fast food industry, for instance, has been criticized for serving foods that have "unhealthy" high-fat content. However, as more consumers demand more nutritious menu items, FAFH opera- tors, to stay competitive, will have to adapt to these nutrition-conscious patrons. Some of the fast food chains have, however, started to respond positively to the demand of the customers for healthier and low-fat food. Except for the intake of calcium and magnesium by certain population groups, the individuals in the sample seem to have generally acquired at least two-¢ age-squared coefficients are not. Less uniformity in the nutrient consumption patterns in relation to the age variable is found using all foods eaten. In particu- lar, a significant positive sign on the age variable and a significant negative sign on the age-squared vari- able is found on protein, dietary fiber, fats, monounsaturated and polyunsaturated fatty acids per 1,000 kilocalories, magnesium, copper,and potassium consumption models. Conversely, a significant negative sign on the age variable and a significant positive sign on the age-squared variable is found on carbohydrate per 1,000 kilocalories, food energy, thiamin, vitamins B6 and Bu, folate, calcium, phospho- rus, iron, and sodium consumption models. Nonlinear consumption patterns are also reflected in some of the significant coefficients of the income and the income-squared variables. Household income is more of a significant factor affecting individual consumption of nutrients from FAFH than from either FAH or all foods eaten as evidenced by the number of significant coefficients. Moreover, the general pattern of nutrient consumption from FAFH typically starts with an increase and then declines with successive positive increments of income, as indicated by the significant positive and negative signs of the income and income-squared coefficients. This consumption pattern is evident in the following consumption models from FAFH: protein, fat, saturated fatty acids, monounsaturated and polyunsaturated fatty acids and dietary fiber per 1,000 kilocalories, food energy, thiamin, riboflavin, niacin, folate, magnesium, iron, and sodium. Consumption models from FAFH with a significant and positive income coefficient but insig- nificant income-squared coefficient are alcohol; vitamins A, E, C, and B6; carotenes; calcium; phospho- rus; copper; and potassium. In contrast, the general pattern of nutrient consumption from FAH starts with a decrease and then increases with successive incre- ments of income. Individual consumption of food energy, cholesterol, riboflavin, niacin, phosphorus, and sodium from FAH depicts this same pattern with increases in household income. In the models for monounsaturated fatty acid per 1,000 kilocalories, phosphorus, iron, zinc, and potassium consumption from FAH, coefficients are negative and significant for the income term but insignificant on the income- squared term. If all the foods eaten are considered, the consumption of carotenes, vitamins E and C, and copper initially increases and then decreases with successive increments in income. The coefficients associated with the income-squared term in the models for polyunsaturated fatty acid and dietary 22 fiber per 1,000 kilocalories, vitamins A and B6, folate, magnesium, and potassium from all foods eaten are not statistically significant although the income ter are positive and significant. In general, these res . contrast to those of the Mndham et al. (1983) study, which used the 1977-78 NFCS data set. Their results indicate that income had no statistically significant effect upon the nutritional quality of the diets. Conclusions Heightened consumer interest in health and nutrition has increased the need for a complete understanding of nutrient consumption patterns not only from food eaten at home but also from food eaten away from home. Moreover, food consumption relationships are traditionally specified between socioeconomic factors and quantity or expenditure measures. These approaches, however, do not allow inferences regarding the nutritional status of an individual's diet. On the other hand, although consid- erable literature exists on demand models for nutri- ents, little attention is paid to the analysis of the demand for nutrients derived from either FAFH or FAH. In this light, this study evaluates individual nutrient consumption to provide a comprehensive Q description and understanding of nutrient consump- tion patterns not only from total food consumption but also from FAFH and FAH. In addition, the infor- mation obtained from this study could be used as a guide by nutrition educators and policy makers in the design and development of nutrition educational programs. The results in this study generally indicate that average nutrient intakes from FAF H are lower than average nutrient intakes from FAH except for fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids. This information has important implications for the away-from-home food industry. The fast food industry, for instance, has been criticized for serving foods that have “unhealthy” high-fat content. However, as more consumers demand more nutritious menu items, FAFH opera- tors, to stay competitive, will have to adapt to these nutrition-conscious patrons. Some of the fast food chains have, however, started to respond positively to the demand of the customers for healthier and low-fat food. Except for the intake of calcium and magnesium by certain population groups, the individuals in the, sample seem to have generally acquired at least twoy thirds of the recommended daily allowances from all Foods, which indicate that these individuals are Fcally consuming adequate diets. Moreover, results _ .trally indicate that distinct differences exist between the type of socioeconomic and demographic characteristics that significantly affect the intake of nutrients from food away from home and food at home. These findings illustrate the importance of socioeconomic and demographic variables in nutrient consumption models. Literature Cited Adrian, j., and R. Daniel, Impact of Socioeconomic Factors on Consumption of Selected Food Nutrients in the United States, American journal of Agricultural Econom- ics, 58(1976):31-38. Akin, ].D., D.K. Guilkey, and B.M. Popkin, The School Lunch Program and Nutrient Intake: A Switching Regression Analysis, American journal of Agricultural Economics, 65(1983):477-85. Basiotis, P., M. Brown, S.R. Johnson, and KJ. Morgan, Nutrient Availability, Food Costs, and Food Stamps, American journal of Agricultural Economics, 65(1983):685- 93. Bethlehem, ].G., Reduction of Nonresponse Bias Through Regression Estimation, journal of Offic. Statistics, . 4(1988):251-260. ’ ‘a, S.T., A Healthy Diet with Animal Product Options: ‘ What the Food Marketer and Consumer are Doing, Food Marketing Institute, Washington, D.C., November 1988. Capps, 0., ]r., and j. Schmitz, A Recognition of Health and Nutrition Factors in Food Demand Analysis, Western journal of Agricultural Economics, 16(1991):21-35. Chavas, j.P., and K.O. Keplinger, Impact of Domestic Food Programs on Nutrient Intake of Low Income Persons in the United States, Southern journal of Agricultural Economics, 15(1983):155-63. Davis, C.G., and P.H. Neenan, The Impact of Food Stamps and Nutrition Education Programs on Food Groups Expenditure and Nutrient Intake of Low Income Households, Southern journal of Agricultural Economics, 11(1979):121-29. Devaney, B., and T. Fraker, The Dietary Impacts of the School Breakfast Program, American journal of Agricul- tural Economics, 71(1989):932-48. Devaney, B., and R. Moffitt, Dietary Effects of the Food I Stamp Program, American journal of Agricultural Economics, 73(1991):202-11 . Draper, N., and H. Smith, Applied Regression Analysis, Second Edition, john Wiley and Sons, Inc., New York, 1981. Guthrie, H.A., Introductory Nutrition. St. Louis: Mosby Co., 1983. Heckman, ].j., The Common Structure of Statistical Models of Truncation, Sample Selection and Limited Depen- ? dent Variables and a Simple Estimator for Such Models, Annals of Economic and Social Measurement, 5(l976):475-492. —i—, Sample Selection Bias as a Specification Error, Econornetrica, 47(1979):153- 161. 23 Huffman, ]., The Influence of Income and Demographic Variables on American Households’ Achievement of Nutritionally Adequate Diets. Unpublished Dissertation, University of California, Davis, 1988. Lane, S., Food Distribution and Food Stamp Program Effects on Nutritional Achievement of Low Income Persons in Kern County, California, American journal of Agricultural Economics, 60(1978):108-16. LeBovit, C., Expenditure for Food Away From Home, National Food Situation, 152(1967):36-38. Martin, E.A., and A.A. Coolidge, Nutrition in Action. New York: Holt, Rinehart, and Winston, 1978. McCracken, V., and j. Brandt, Household Consumption of Food Away From Home: Total Expenditure and by Type of Food Facility, American journal of Agricultural Economics, 69(May 1987) :274-284. National Food Review, 1990 Edition, ERS-U.S. Department of Agriculture 13(1990) :25. National Research Council, Food and Nutrition Board. 1989. Recommended Dietary Allowances, 10th ed., Washington, DC: National Academy Press, 284 pp. National Restaurant Association, 1988. 1988 Gallup Poll, Washington, D.C.. Noah, T., Married Couples Are Found to Head Fewer Households, Wall Street journal, june 11, 1991, p. A6. Pao, E., S. Mickle, and M. Burk, One-day and 3-day Nutri- ent Intakes by Individuals - Nationwide Food Con- sumption Survey Findings, Spring 1977, journal of the American Dietetic Association, 85(March 1985):313-24. Phlips, L., Applied Consumption Analysis. Amsterdam: North- Holland Publishing Co., 1983. Pindyck, R.S., and D.L. Rubinfeld, Econometric Models and Economic Forecasts, 3rd ed., McGraw-Hill, 1991. Price, D.W., D.A. West, G.E. Schier, and D.Z. Price, Food Delivery Programs and Other Factors Affecting Nutrient Intake of Children, American journal of Agricultural Economics 60(1978):609-18. Putnam, 1.1., and M.G. Van Dress, Changes Ahead for Eating Out, National Food Review 26(1984):15-17. Robinson, C.H., Fundamentals of Normal Nutrition. New York: Macmillan Publishing Co., Inc., 1973. Scearce, W.K., and R.B. Jensen, Food Stamp Program Effects on Availability of Food Nutrients for Low Income Families in the Southern Region of the United States, Southern journal of Agricultural Economics, 11(1979):113- 20. U.S. Statistical Abstracts, 1988-89. Varian, H. Microeconomic Analysis. New York: Norton, 1978. Waldrop j., and T. Exter, The Legacy of the 1980s, American Demographics, 13(March 1991):32-38. Windham, C., B. Wyse, and R. Hansen, Nutrient Density of Diets in the USDA Nationwide Food Consumption Survey, 1977-1978: II. Adequacy of Nutrient Density Consumption Practices, journal of the American Dietetic Association, 82(Ianuary 1983):34-43. Windham, C., B. Wyse, R. Hansen, and R. Hurst, Nutrient Density of Diets in the USDA Nationwide Food Consumption Survey, 1977-1978: I. Impact of Socioeco- nomic Status on Dietary Density, journal of the American Dietetic Association, 82(]anuary 1983):28-34. [Blank Pagé in Orifinfl Bulletin] l *1. i’ 1 ,‘..~ ‘. ,4, 1 n APPENDIX A Means of Average Daily Nutrient Intake for Individuals Across Population Groups w>.¢ w>.¢ ¢@.¢ >~.¢ h>.¢ ¢w.¢ m>.¢ m>.¢ ¢h.¢ m>.¢ >>.¢ m¢.¢ mm.¢ >¢.¢ m¢.¢ ow.¢ >>.¢ mm.¢ >m.¢ ~m.¢ wm.v m>.¢ wm.¢ mm.¢ wm.¢ @m.¢ mw.¢ @m.¢ m¢.¢ mo.m om.v ¢m.¢ mm.¢ mm.¢ mm.¢ ¢¢.¢ mb.m mm.¢ mm.¢ w>.¢ Q>.¢ NN.m >H.m mo.m ~m.¢ mw.¢ AUX o¢¢~\mz¢mwv mmm mmm.¢ >m.¢ mm.¢ m¢.¢ @m.¢ mw.¢ Hm.¢ om.¢ m@.¢ Hm.¢ mw.¢ ¢m.¢ ¢>.v mm.¢ wm.¢ ~m.¢ Hm.¢ mw.¢ >o.m mm.¢ oo.m ww.¢ ~>.¢ AQ4 m~.mH ¢N.mH a¢.mH mm.~H >~.mH Hv.mH m~.mH om.ma ¢a.mH m¢.mH m~.mH mm.NH mm.mH mo.¢H mo.mH mN.mH wH.mH HH.mH ~w.mH m~.mH ¢w.mH w~.~H am.NH om.ma om.mH ww.mH ¢m.mH mm.mH wa.NH No.vH Hm.mH mo.¢H a>.mH >@.mH mw.mH wH.mH Nm.mH >o.mH Ho.¢H mm.mH m>.mH Hm.¢H H>.mH wN.¢H om.ma mm.mH AUM QoQH\mz¢mwV 24h mum» Hfim >w.mH >m.mH m>.mH Nm.NH mm.mH m¢.~H m¢.mH ¢>.mH om.ma om.ma mm.ma ~m.mH om.mH ~>.m~ w~.mH mm.ma H¢.mH ¢m.mH wm.~H mm.mH mm.mH mw.ma >N.mH AA< mH.¢a mm.mH mm.ma ¢m.mH Q¢.¢H ¢m.¢H ma.¢H mm.mH vH.¢a mm.mH Hc.¢H ¢m.¢H hm.¢H mm.¢H @H.¢H oo.¢a Hm.¢H ¢m.mH wm.mH ¢H.¢a mo.va mm.mH ¢~.vH >m.~H >o.mH mh.NH ~m.¢H mm.~H ww.~H wH.~H ¢m.NH aH.mH m>.~H om.NH Nm.¢H m¢.HH w~.HH ww.~H oo.mH NH.NH ¢Q.mH mo.mH m¢.mH N>.Na aH.ma ¢b.NH ADM Qo¢H\mz.mH wH.mH om.mH ¢>.¢H aH.¢a mm.mH No.vH wm.mH Hm.mH mo.mH mm.¢H wm.mH ha.vH mm.mH flm.mH Hm.~H m@.mH oa.¢H mm.ma mm.mH mH.¢H Aq< >m.mmmH mQ.~mm mH.¢w~H mw.m¢w @¢.mwmH ~>.H~w >m.@¢~H @¢.mH¢ ~@.m¢mH mm.HQw ¢m.H¢mH Hm.mm¢ ~m.¢>mH wm.mw¢ mm.¢~mH H>.w¢w >¢.mmHH >¢.~>¢ ¢m.HmmH m~.m~> mm.mmmH m~.o¢w m¢.mwmH mm.Hom mm.mmmH mw.omm oH.~mmH >H.om¢ w¢.mQmH mw.mow ¢Q.Q¢mH ¢m.mmm @H.om~H -.¢>m o>.m>~H mm.@~w ~H.~HmH Hm.mmm H¢.mHmH >>.>wm mw.~@mH w>.wmm H¢.>¢mH ~m.mom wm.~@~H wH.m>m AwmHmoq¢H Nm.a>hH mm.mmma am.HNmH wm.mmmH NH.>m¢H >H.¢@¢~ ~w.>m>H H¢.m¢mH mw.mNma ~w.wm>H mm.~>wH Qw.m@>H HH.m¢>a mm.~H@H @~.m¢>H mm.mo>a >m.mm>H m~.~m>H ~m.m>wH AA4 >mvxwmz wcmxomz MEWS OZ mm» umfln Hmfiummm uoz mcw>@mowm mEmuw Uoom mew uoz ©m>0HmEm u:mE>oHmEm UHMEUM mam: wow mmfincoz uflcmmmfim cfimfiwo uuzuo umm\cmfim4 xumfim wufinz wumm umwz nusom uwmsuflz ummmsuuoz coflmom 0numE|qoz cmnusnzm cmnuo GOHUMNHGMDHD m®HQMHHM> 0HnmmnmoEm0 26 mN.mm mm.mm ~¢.mm ~m.>m bo.mm mm.mm mm.o¢ om.mm mm.mm mo.mm >~.mm mo.N¢ wm.mm mm.hm mm.mm >m.mm mo.o¢ oN.mm am.mm mH.mm am.mm wH.mm oa.mm mw.>m ~>.mm ww.H¢ mm.wm mm.wm om.mm wm.o¢ H¢.mm H@.>m m¢.mm ¢m.mm mm.mm om.mm HH.mm Nm.wm Ho.mm mm.>m wm.mm ~H.mm mH.mm Ho.mm mH.o¢ mm.mm ADM QQ¢H\w:mommm wH.mm mm.rm ~¢.@~ mm.hm >H.@m wH.wm mw.mm mm.mm mm.wm ¢w.wm Hm.>m ¢H.mm wN.mm mN.mm mm.wm AA4 mw.¢- >H.om mH.¢m~ HH.@m ~w.@- @¢.mm wm.QmH @m.mw mH.m- Hm.~m >m.Hm~ ¢m.w> w¢.¢m~ mm.m> wo.mH~ -.@m wH.omH mm.¢> ~>.~w~ ww.m¢H mm.H- @~.~m mm.H~m ¢m.¢@ ~H.m~m m@.m> wm.~@~ ~>.mw ~¢.Hw~ ~>.wm Q¢.wH~ m¢.~m mm.Hm~ ¢m.¢@ m¢.H¢~ om.@m o>.¢Q~ @>.@@ HH.mHm mm.~m ¢>.H¢~ mQ.mm m>.mH~ HH.mm Hm.mH~ >m.>@ AmZ4mwHAAHZV m>.>¢m m>.mm~ ~H.m¢~ oa.mmN ao.mwN >¢.¢hN wo.>m~ om.mmN mN.m¢m mm.mmN >m.wmm o¢.mmm mm.wNm mw.mHm w¢.~@~ mo.omN mN.mom mm.mwN mm.mhN mm.mmN ¢H.@m~ ¢m.@>~ mm.N w¢.~ Hm.N ~m.~ fim.~ om.N w¢.~ ~m.~ ~m.~ Hm.N ~m.~ mm.N m¢.~ ma.» 2a ~m.~ m¢.N m¢.~ wm.N Hm.~ b¢.N wm.N m¢.m o>.N ¢w.~ mw.N ad.» o>.~ ~Q.~ mm.N o>.N mm.N mm.N >m.~ ¢m.~ mm.N mN.N N¢.N ah.N ~w.~ >>.~ ~>.~ ¢m.~ >>.N ww.~ ~w.~ Aux Q¢¢H\wzmmwv mmm mmw.~ ow.N m¢.~ mm.m mm.N Hm.~ >m.~ ow.N Nm.N ~w.~ mo.m >¢.N Hm.~ >m.~ Hw.~ Nm.N >m.~ >m.~ mw.~ vm.N AQ4 ¢m.¢ mm.¢ mm.¢ mm.¢ ~m.¢ hH.m ¢m.¢ mm.¢ mw.¢ mo.m mm.¢ >m.¢ oN.m mm.» mo.m Hm.¢ mm.¢ >m.¢ mo.m ¢@.¢ mH.m Hm.¢ m>.¢ mH.m ~H.m wH.m mo.m mH.m ah.¢ mN.m ¢H.m oN.m mH.m >H.m mo.m w>.¢ mm.¢ mo.m ma.m wm.¢ mN.m ~¢.m mm.¢ wm.m mH.m mo.m RUM ¢¢¢H\mz¢mwv m¢m mmmm MQHU4 amm OZOZ mo.m >¢.m ¢H.m m>.¢ wo.m ma.m ¢o.m ~H.m mm.¢ H~.m mo.m Ha.m mN.m oN.m ¢H.m mo.m >m.¢ mH.m oN.m wm.¢ >~.m ho.m ¢m.¢ QA< >m©xwwz vzuxmmz MMMB OZ mo» uwfia Hmfiuomw uoz m¢fi>fi@u~m mémum voom méw uoz vm>0HmEm ucoehoamsm mfimsom wfimz xwm mmfisnoz oficmmmfim Gflmwno Mwnuo omm\¢mHm4 xomflm mafia: oomm umwz nusow umwsvflz ummmnuuoz Gowmmm ouuwZ|noz CMQHBQDM cmnub cofiumuwcmnub MUHDMHHM> 0H£mmumoEon A, 27 @>.mwm >H.@@~ ho.Hmm mm.H¢¢ mm.Hmm mm.NmN ww.~wm ~>.mmm >>.wmm mH.mmm wH.m¢m ma.wmn mo.mov Nw.mNm m>.¢¢~ ¢m.mmm w>.Hmm >>.m¢m >m.¢¢~ >m.mwm mm.mmN ¢w.¢mm mw.mhm ¢~.moH wm.~m wo.mm >H.m- oH.HoH oN.woa ¢H.~¢H mm.ooa >¢.¢¢H wo.Noa mm.ooa @m.mHH mm.moa mm.>m w~.¢m @m.H¢H @m.¢¢H m~.¢m mw.Nm ¢@.mHH ~@.»m >m.ooH ww.>¢H AWM m:4muomuH:. mam mmmm mmzmaom¢u Hm.mm¢ mo.mmm mw.mmm m¢.H~m ~H.-¢ ¢o.m>~ m~.wH¢ wo.mH¢ mm.No¢ ¢¢.m~¢ ~m.mH¢ wm.oNv wm.w¢¢ om.wmm m¢.oom H¢.m~¢ mw.~w¢ mv.mmm No.aH¢ Hw.Hm¢ m>.owm m>.>~¢ m@.Hm¢ QA4 o¢.Nwmm Hm.mmm¢ mm.mmmv N¢.mmmm H~.HmHm Nm.Nmo¢ ¢m.oNvm mm.¢>@¢ w~.~m>¢ wm.Homm Nm.mmom m~.Hm>¢ ¢¢.HNNm mo.m¢mm Hm.@¢mm mm.oNNm mm.mwvm >@.mmw¢ m>.H¢Hm ¢o.o>~m @@.@>w¢ >~.~oHm mm.Hmvm ~¢.@H¢H m¢.w>~H wm.mmmH >w.m>¢H w@.~>mH H~.¢w~H oa.mNma cm.wmmH vm.mmma >o.wQ¢H ¢m.H>mH >¢.~mmH wm.mmNa mm.mmHa ~m.m>H~ ¢@.mmmH om.mmma wm.~@~H mo.NNma wm.Nmma mo.mHmH mm.mmma ¢w.¢m¢H AwHHZD .mWHZHv mmm mm >m.wm~w mm.owmm ¢m.aomm mm.¢omw wm.wmow m>.m~w¢ HQ.mHHw wm.¢wmm No.N>wm ~>.H¢¢m mN.mmow Hm.homm wH.m¢>m mw.~m>¢ Hm.>~w¢ m@.>H~w mm.mo¢m wH.w¢mm ¢~.mm¢m mm.Hmmw mm.Nomm ¢w.Nmom mm.mwww JAG mm.o Hm.o mm.o mm.o >m.o mN.o Hm.o ¢m.o ov.o Nm.o ww.o mm.o -.¢ ¢>.¢ ¢@.¢ ¢w.o >m.o Hh.o H¢.o Nm.o >~.¢ ¢@.o Nm.o Nm.o m~.H Nm.o mm.o Nm.o mo.a wm.o Hm.o mw.o NN.H mm.o mh.o ~>.o mm.N @¢.¢ >m.o NN.H ~>.o >m.o mm.o mm.o Ho.a ma.H AUM oQ¢H\mz¢mo. mmm mmmm AOQOUA4 om.o mm.o a>.o Hm.o o>.o mm.o mm.o mm.o m¢.o mm.o mw.o mm.o oN.o mm.H mm.o oh.o mo.o Nm.o wm.o wh.o ¢m.o Nm.o w>.Q AA4 mw.N >¢.N m¢.~ mH.m mm.N mN.N mw.m mm.N mm.N ~m.~ >m.~ mm.N >h.N @m.H mH.~ mw.N o>.N ~w.~ Hm.N w¢.~ ¢¢.N mw.N mm.N Nm.H om.H >m.H mN.N Nm.H ww.H ¢o.~ hm.H oo.N ~m.H Hm.a No.N m@.H om.H ~>.H ¢m.H ¢o.~ >@.H mm.H mm.H >m.H @@.H ¢m.H AUM Q¢¢H\wz<¢oV mmm mmmm mmmHm »mm.~ m¢.~ m>.~ @w.~ ¢o.~ ¢H.~ m¢.~ Qm.~ ¢¢.~ ¢¢.~ mm.N mm.N @¢.~ m¢.~ QA4 xmvxomz ucmxuoz MWMS OZ mow uwfln Hmfiuumm uoz mcw>flmomm msmum doom mew uoz vo>oHmEm ucoE>oHmEm mflmemm mam: xum mmflnnoz owcmmmfim ufimfluo umnuo omm\cmHm¢ xumam uuflnz oomm um»: nusom umwzwfiz ummmsunoz cowmmm 0uuwZ|coz cmnusndm cmnub cofiumuflcmnub WUAQMHHM> uwnmmumosoa 28 >m.H mm.a mm.H mm.H @m.H am.H m¢.H Nm.H oN.H wm.H >m.H mm.H ¢¢.a Na.H HN.H o¢.H m¢.H mm.H wm.H Hm.H ¢¢.a wm.H Nm.H mv.o ¢¢.o m¢.o Nm.o ¢¢.¢ wm.o mm.o w¢.o mm.o Nm.o ¢¢.¢ >m.o o¢.o ¢m.o H¢.¢ ¢¢.o ~¢.o ¢¢.o ¢¢.¢ m¢.o m¢.o ¢¢.¢ H¢.o AwZ4MOHAQH2V m4AmOmHM ww.H Nh.a oh.H mm.H mw.H h¢.a mw.H mm.H ¢¢.~ ¢m.H mm.a ¢w.H Hw.a mm.H m¢.H Hh.H Nh.H mw.H oh.H Hw.a Nh.a mw.H Hm.H qA< ho.H w¢.H mo.H ¢o.H >o.H mo.H mH.H ¢o.H mm.o m~.H >o.H wa.H >H.H mH.H H¢.H >o.H mo.H HH.a >¢.H mo.H NH.H >o.H H¢.H wm.o wm.o >m.¢ mN.o mm.o om.o mN.o mm.o mN.o m¢.o wm.o om.o am.o ¢~.o mm.o mm.o ¢m.o mm.o mm.o mm.o mm.o rm.o mm.o Awzm.H m~.~ mm.a hN.H hm.H mm.H mN.H mm.H ¢m.H m~.H AA4 mH.>> -.¢~ No.H> No.HN m>.~> om.oN m>.~m ¢m.>H ¢Q.w> mm.oN mm.Nw >N.wH oH.Ho Nb.wH w¢.~> om.aN mm.o> >m.>H Hm.om mm.mN ao.m> @¢.o~ H>.m@ Nm.oN m¢.m@ ¢¢.wH ¢m.~m mm.¢H ho.om wH.@~ m~.¢> w>.mH ¢H.>> >¢.- wH.Nh mm.wH Nm.H> mm.mH HH.m@ mm.NN m¢.ww wm.mH mw.m> mm.oN o@.>> H>.oN AmE€dUHAQHZV mmm mm m¢.om mo.rm mm.>w ~>.mo om.om mm.o> mh.mm mN.mm ww.Hm ~w.>m NN.mm co.hm NN.mm Hm.mm om.wm Nm.mm mw.Nm Ho.mm ~m.¢w ¢w.@m mw.mh ¢>.mm mm.Hm H QQ< oa.m ¢H.w oa.m H~.w mH.w w¢.m wm.m No.m >~.m ¢o.> mo.» mm.w >m.w o¢.w >m.¢ ¢~.w mH.w oa.m ¢Q.w ¢H.m mm.m m~.w mm.m m¢.~ >m.~ om.N ¢H.~ w¢.~ ww.H ww.H >w.~ mo.N Hm.N wv.~ wo.N >¢.~ ww.H ¢N.N mv.~ w¢.~ >w.~ mm.N HN.N mm.N mm.N ~m.~ Amem m:¢muHqqH:v m m>.> wm.> mm.» mm.» ~m.> mH.w >N.h ¢H.m ww.w Ha.m ~m.> oo.m m>.> w>.> Hm.w No.m mm.» mm.» ¢h.h mm.» Hm.» NH.m ¢¢.> Aqfi >m@xmw3 UGUMUUZ xmmz OZ mu» umfla Hmfluwmm uoz mcH>wwuwm memuw noom mEu uoz U0>0HQEW uGUE>OHmEm mfimemm mfimz xum mmflncoz uwcmmmfim nflmfiuo Hwnuo omm\cmfim< gumfim wufinz wumm um»: nusom umwawflz ummosuuoz cowmmm ouum2|:oz QMDHDQUW cmnub COHUMNHCMDHD MUHQMHHM> uHnmmnmoEmQ \ 29 m¢.¢ ¢m.m ¢¢.¢ mw.m m~.¢ mw.¢ wm.¢ mH.¢ mm.» wm.¢ Hm.¢ mo.¢ ¢@.m >m.~ mm.¢ ~m.¢ mo.v m~.¢ >a.¢ m>.¢ w~.¢ @~.¢ m¢.¢ wm.H mm.H w>.H mo.H mw.H mm.H >m.H ¢w.H m~.H >o.N mw.H om.a mm.H coma 0N4. m>.H mm.H vm.H wm.H ~m.~ m¢.H mm.H Hm.H Amz m¢.m m¢.m mw.m mm.¢ m¢.m mN.m m~.m mm.m mm.¢ om.w om.m m>.¢ o¢.¢ om.m mo.m wm.m No.m Nm.m Hm.m ¢m.w HN.m ow.m m¢.m HA4 @m.wwH wm.mm NN.m>H mo.mm wH.~wH NN.mm mo.mmH mm.~¢ mH.mmH mo.mm ~w.¢>H >m.m¢ mN.mmH H~.m¢ w¢.w>H mm.wm mm.Nma mm.m¢ ¢H.@Q~ ¢m.mw m~.mwH m>.¢m ¢¢.o- w>.m¢ ~¢.wH~ mm.Hm >w.mmH ¢¢.m¢ mH.¢wH vw.mm wH.wwH mm.¢m >~.¢mH mm.mm @H.mmH mH.>m @w.~>H >a.Hm w>.Q@H ¢>.mm ¢m.~mH Hm.Hm mH.>mH mm.mm m¢.m>H mm.mm Amz.Q- mm.NNN mo.mma Hm.¢m~ hm.HNN m¢.>¢~ mm.mmN ~>.@@H Hw.@mH mm.vNN HN.NNN mm.¢m~ o¢.moN mN.>HN @H.mH~ Hm.>- ma.hHN Qq< mN.H oN.a NN.H w~.~ m~.H Na.a @~.H mH.H mo.H a¢.H NN.a >m.H w~.H mo=H mo.H ¢N.a mN.H mN.H oN.H m~.H oN.fi m~.H ma.H mm.o o¢.o o¢.o Hm.o mm.o mm.o Nm.o ~¢.o Hm.o >¢.¢ mm.o wm.o >m.o mm.o ov.o mm.o wm.o H¢.Q mm.o mm.o mm.o o¢.o mm.o AmE m¢.H Hm.H om.a w¢.H Hm.H bN.H m¢.H Nm.H wN.H w>.H m¢.H mm.H Nm.H ¢m.H om.a Nm.H m¢.H Hm.a m¢.H m¢.H ¢¢.a ¢m.H m¢.H AA4 mH.ma ¢>.m mm.¢H mm.m mo.mH mm.m m@.¢~ wm.¢ ¢o.mH mm.m mm.mH m@.¢ @¢.mH >m.¢ Hw.¢H mN.m wm.- Hw.¢ ¢m.>H mo.> @m.¢H mm.m HN.ha H¢.m ~m.wH mm.m m¢.¢H mw.¢ wm.¢H NH.m mo.ma om.m mm.¢H m¢.m ¢¢.mH mH.w ~>.¢H Hw.m No.ma mm.m w~.mH mm.m ~H.mH No.w Hw.¢H Hw.m Am24MUHAQHEv mmm mmmm ZHUQHZ mm.wH ~¢.mH Hm.mH w>.>H oN.mH m¢.>H @w.>H Hm.mH mo.wH om.NN >o.mH mN.oN m~.mH No.wa Ho.mH w~.mH mw.mH @w.m~ mw.mH m@.wH ~m.wH Hm.mH ¢¢.mH AJ< >mwxmoz ncwxmmz mmmz OZ mm» umfla Hmfluwmm uoz mGH>flm0mm memum Uoom mEm uoz ©w>oHmEm uGOE>OH@EM mflméwm mfimz xww mwflncoz uficmmmflm cfimfiho umsuo umm\cmwm< xumflm wufinz wumm umwz sunom umwsnfiz uwmwnunoz cofimwm onumZ|c0z GMQHDQDW smnnb COHUMNHGMQHD MUHQMHHM> oH£mmnmoEmQ 30 mm.m mm.m m~.~H Hm.mmH mm.¢w ~m.~m~ ~m.w>w mm.Hmm oo.mmoa >¢.mmm HN.>ma H>.omw >mU¥ww3 wm.m mm.m Nm.NH wH.w@H mm.mw ¢m.wmm ¢H.¢@@ mo.m¢m om.m¢HH m>.Nmm >m.mma >m.mm> uawxwws MMMB mm.m wm.m @¢.~H >m.>mH mm.mw >¢.¢m~ mm.mmm Hm.>¢m ww.@~HH mo.mwm -.>Q~ ~@.¢H> OZ ¢>.m ~m.~ N¢.HH mm.mmH mm.om Hm.omm mm.m¢m m~.H¢~ oH.HooH m>.@¢m mm.mma mm.Hmw mm» . umfin Hmfiuwmm >m.m >¢.m m¢.NH >m.mmH mw.mm om.mmN Ho.amm >m.mmm Hm.oNHa Hw.o>m om.mmH mo.mH> uoz mm.m am.m mm.oH wo.m>a mw.m¢ wm.¢mH H¢.¢¢m mm.m>N Nm.mmm mo.Hm¢ mm.wmH mm.mmm mflH>Hw0wm memum Uoom mm.oH >>.~ ¢¢.~H ma.wma mm.mm mm.mmN mm.>om HH.mw~ om.w¢oa ¢H.mwm ¢H.¢wH mo.N>m mew uoz mm.m o>.m ~m.~H >m.¢mH mm.mw am.mmm ¢m.mwm m¢.Hwm mm.m¢aH mm.mmm m¢.~H~ ~H.-> Um>0HQEm ucwE>oHmEm mw.m m>.~ >¢.oH >H.>wa w>.~m ¢m.Hom ~>.ww> Ho.mmm mm.m¢m mo.Nom mm.mmH om.wom mfiwemm H¢.aH HH.¢ m¢.¢H ¢w.~H~ m¢.>> ¢m.mwN mm.NooH mm.mo¢ mm.mmNH wN.>mw mm.>mm mm.aam wfimz xmm mm.m wv.m Nm.NH ow.>mH mm.mm m¢.mmN m>.¢>@ wm.mmm ¢m.oHHH mm.>wm Nm.mmH ¢w.w¢> mmflncoz om.HH am.m -.mH m>.mHm ma.mm H¢.Hm~ mo.¢HoH mm.mmm mm.H>HH m¢.mmm ¢m.NmH mm.oNm uficmmmfim cwmwuo mm.HH mm.N m>.~H ¢¢.mH~ mN.mm ¢H.N¢m Hm.mooa mm.mHm mm.>¢HH >m.mwm HN.mmH >m.~mw nmsuo ~¢.m ah.N mm.HH ¢m.N>H mm.wm oH.¢HN >m.wom mv.>m~ Nm.¢mm mm.mm¢ >>.¢wH ~>.mmm umm\GmHm4 mo.m om.m oa.HH mo.mma wm.mm m>.mmH wo.H@> >w.¢~m om.¢>m wm.am¢ mm.~>H mm.mmm xumfim No.oH w¢.m om.Na ~m.~mH mm.mm mm.mmN mo.mmm >~.>mm mm.m~HH mm.¢mm mm.HoN ww.m~> wufiss womm om.m mm.» o~.~H w¢.oom mH.ww mw.m¢N mo.Nmm mm.mHm o>.m¢HH mN.mmw >h.amH @H.ww> uwwz ~¢.¢H mm.m mm.NH Hw.>mH wm.ww m>.mmm wm.>>@ mm.m¢m m>.mHH~ o¢.vmm mm.mma mm.mmm susom Nm.m om.m ¢m.HH mm.~@H mm.¢m ¢w.m- mo.mwm >m.>mm mm.moHa mm.>>m @@.¢o~ ¢m.¢~> um03©wZ m>.m om.m NN.NH m¢.wwH >~.Hw ¢~.m- Hw.m¢m mN.wNm >m.N>oH mm.o¢m HN.ooN HN.h>w uwmwsuuoz cowmwm NH.oa ‘ Hm.» NN.Na w¢.@@H Hw.~w N~.w- mw.¢om w>.omm >h.mHaa m¢.mwm mH.>mH Hw.mH> o~»mz-¢oz ao.oH >m.m am.NH mm.HmH w~.ww Hm.mmN ¢~.@>m ~¢.~¢m mw.>~HH ¢m.mmm ¢m.moN ¢m.¢H> GMQHSQSW mm.m mm.m ¢m.Ha mw.~wH mw.¢w oo.>Nm om.m¢w aw.mHm >~.@woH H¢.¢¢m ¢~.mwH Nv.mwm GMQHD cmwwmuwcmnub AWZAMUHQAHZV AmE¢MOHAAH2v Aw2¢MUHAQHZv AwE4MUHAAHZV mmm mmmm AA< mmm mmmm AA4 m zomH :aHwmzwmz wnmommwomm ZDHUA4U ufinmwumoewa 31 f Q= l“ mw.NmmH mm.mmw w~.mm@H mN.mmm mw.NmmH ~m.¢H> Hm.maoN ¢H.mwm ¢m.>HmH ¢¢.~Q> wo.mm>a w¢.@mm ¢m.>ooN mm.m>m Hm.~mmH mm.m¢> mw.mmma ¢~.~wm mH.¢¢H~ om.Hmw wm.oomH ¢H.mmm H~.mwH~ Nw.oam wm.¢>H~ ¢~.¢mw wo.>mwH mm.mw¢ No.omma Ho.wmw mw.m¢mH ¢~.~o> ¢m.~wmH mm.mbw ~m.~>wH Hw.¢o> ¢m.HmwH >H.¢H> ~m.wmmH Nm.mmw mm.mmmH mm.mmw >H.mmmH Q>.mQ> ~>.wo@H wm.m>m AmEmm~ ma.wmmN om.mbmN mw.wH¢~ mm.mwmH m>.momN o~.~¢¢~ Hm.o>oN wH.Qm>~ ¢m.mwm~ mm.mNmN Hm.mw¢~ NN.oHoN ~m.m¢mH ¢w.Hm¢~ >o.om¢~ Nm.owmm H>.Ho¢~ wm.¢Q¢~ hm.ommN >~.>¢¢~ ~m.m>- AA4 mm.@@- ~m.~mm ¢Q.~m- @m.HoQH mm.¢-~ ~w.m~QH ww.wmmH >>.>¢> ~m.w>- w>.HQoH -.mHm~ mH.H¢m mm.mHm~ H>.>m> Hm.¢m- mm.H>OH HH.w¢mH mm.mHm ww.~m>~ ¢m.m>HH mH.¢@- ~H.~ooH >m.mH>~ m>.H>> mm.¢@w~ >m.omm mm.mm¢~ >m.~m> w~.>m-. mw.mwm ~H.mw- Hm.mQQH m@.mH- QQ.>mm wH.H¢m~ ~w.mmoH wm.wm- @m.wmoH ~w.HH- mm.~¢m wH.@@- ~>.HfioH Q~.m@- >¢.oHoH ~m.wwH~ ¢H.Q¢m Aw:¢~ mm.ommN mH.m>w~ >w.o¢>~ ~>.w¢Hm wo.mmmN mm.mHwm m¢.mwmm mo.omHm wm.Nmom Nm.moom ¢~.mH@~ >m.Hmm~ mm.mwmN m¢.mmom mm.m¢om >m.mmwN mm.mmom mm.moom wm.oHmN QA< mm.o om.o mm.o mw.o mm.o ww.o mw.o Hw.¢ m>.o ¢m.o mm.o mm.o mm.o mb.o w>.o mm.o mm.o mm.o om.o ¢@.¢ mm.o vm.o mm.o am.o mm.o mm.o ¢N.o Nm.o mN.o wN.o ¢m.o hN.o >m.Q Nm.o mN.o >~.o mN.o Nm.o Nm.o Hm.o mm.o H~.o Hm.o mN.o mm.o Nm.o AWSAMUHAAHZV mmm mmmm mmmmou mo.H mo.H wo.a oo.H wo.H >m.o Ho.H mo.H Nm.o aN.H mo.H NH.a >o.H wm.o bm.o >o.H mo.H mo.H mo.a mo.a ~o.H mo.H mo.H AA4 @m.> mm.> mo.m Gm.» wm.> >m.> om.m ¢@.> >>.w Hm.m ~m.> Hm.m am.m HH.b Hh.> >m.> wH.m mN.m >h.> mw.> om.m mm.> ~@.> @H.m Hm.m ~m.m m¢.~ mN.m Ho.m >w.~ m¢.m Hm.~ mm.m mN.m mm.N Hw.N m¢.~ m¢.m HN.m Hm.N om.m mm.» wo.m mH.m wN.m ma.m AmE4MOHAAHZv mmm mm.m mm.oa ¢¢.@ oN.NH om.oa mN.Ha ~¢.HH ~m.@ w>.m >N.oH mH.oH mw.oH wo.oa m>.m ¢~.¢H >N.oH mm.m QA4 >mUxmw3 vcwxmuz MWWZ OZ mww uwwn Hmfiommw uoz mcw>flo0mm memum uoom mew uoz ©m>oHmEm ucmE>0HmEm wfimemm mam: xmm mmfinfloz uficmmmfim =HmH»o nmsuo 0mm\cmHm< gomfim wuflsz momm ummz susom ummsuflz ummmsunoz cofimwm ouumE|noz GMQHSQDM cmnub GOMUMNHGMDHD mUHQMHHM> 0fi£mmum0EmQ 32 APPENDIX B Means of Nutrient Intakes as a Percentage of the RDA 33 Table B.1. Means of nutrient intakes as a percentage of the RDA - by race. Others Nutrient White Black Asian/Paca Food Energy ALL 73.38 69.32 70.09 72.65 FAFH 24.09 24.57 19.14 22.58 FAH 56.08 54.62 56.16 62.23 Protein ALL 128.99 127.41 130.29 140.07 FAFH 41.3 42.78 31.88 37.76 FAH 99.35 101.94 107.1 123.44 Vitamin A ALL 139.19 105.07 109.46 127.23 FAFH 31.26 26.49 26.01 29.64 FAH 116.92 89.79 90.54 114.93 Vitamin E ALL 88.69 70.88 88.39 83.99 FAFH 27.38 25.21 20.89 26.96 FAH 69.05 55.81 73.19 71.43 Vitamin C ALL 146.61 160.49 155.85 156.86 FAFH 32.86 46.89 24.44 27.33 FAH 123.19 132.97 138.08 147.25 Thiamin ALL 108.36 101.76 109.18 101.33 FAFH 28.47 31.67 19.64 24.44 FAH 87.99 83.05 94.89 91.04 Riboflavin ALL 119.08 101.65 95.1 107.75 FAFH 29.98 28.43 23.26 26.78 FAH 97.66 85.04 78.17 96.32 Niacin ALL 119.91 112.18 108.19 114.23 FAFH 35.18 37.16 28.05 31.47 FAH 94.70 90.09 87.79 100.27 . i” 1’ Table B.1 Cont. Nutrient White Black Asian/Pac” Others Vitamin B6 ALL 84.33 73.52 74.98 82.97 FAFH 21.61 22.40 19.78 20.97 FAH 68.88 60.32 60.59 73.97 Folate ALL 117.01 104.21 99.08 119.29 FAFH 28.25 30.68 23.66 26.90 FAH 96.84 86.18 81.86 108.29 Vitamin B12 ALL 278.6 253.85 165.36 218.31 FAFH 87.72 60.05 50.03 61.78 FAH 215.66 219.50 128.97 190.65 Calcium 7 ALL 83.87 60.04 62.22 67.55 FAFH 22.43 18.95 18.31 17.93 FAH 67.81 48.82 48.89 59.70 Phosphorus ALL 130.93 110.82 113.76 121.63 FAFH 37.98 36.18 28.81 31.60 FAH 103.72 89.35 92.81 107.91 Magnesium ALL 75.31 60.24 68.76 74.96 FAFH 20.31 19.99 17.72 20.08 FAH 60.78 48.36 55.87 66.14 Iron ALL 108.93 93.8 91.66 105.19 FAFH 30.05 29.17 21.27 24.64 FAH 87.42 76.56 76.19 94.95 Zinc ALL 75.45 72.92 67.37 79.81 FAFH 23.42 25.55 18.99 20.52 FAH 58.65 57.66 53.55 70.94 “Refers to Asians and Pacific Islanders. 35 mm.¢m mo.mm ¢¢.Nh wm.¢w mm.¢m mH.>m mm.w> ww.>w Hm.~m ¢m.m> mmm wm.¢H om.ma ¢H.m~ H¢.bN m~.>~ mm.mN Nm.mN wo.mN mo.¢m mH.m~ mm4m m ZH24HH> mm.Nw om.>w No.mw mH.¢m H¢.m> mo.wb mH.Hm >N.om >m.om mm.bm AA< Hm.Nm mm.hoH N¢.¢¢H am.mHH >¢.~m oh.mm bm.bNH ~m.¢oH ¢~.>> >H.wm mmm mm.vm m>.wH mm.wm NN.mm mw.¢~ m>.m~ m¢.mm ¢>.w~ mm.¢~ >~.m~ mmmm 4 ZHZ4HH> ~G®EO3 ucmcmwnm now "wwdHm> mcfiumfisoflmo now mumz >n uwm: mum om.oaH mm.oNH mm.¢wH wm.mmH >m.moa >m.wHH ww.>¢H >m.m~H >N.@m mm.H~H AA< >m.ooa H~.mm mH.NoH m~.¢m Hm.om mm.moa om.¢oa oo.¢oH mm.mm ¢w.w¢H mam .mnunoE w uwuflw onu .cmEo3 mcflumuumfi now wH.m~ ~¢.mH oo.¢m Ho.wm ~>.@¢ om.m¢ ¢>.mm m~.~¢ mm.Nm mm.Nm mmmH mm.omH ~¢.~mH Qqfl wo.wm ~>.¢m mm.Nw ov.Hm >H.>v ~w.¢m wo.mm m>.~m Hm.m¢ w>.mm NHUUMUEHHU UHHSU USU mnfiaofifiom wne "wuoz¢ HQ.>H ~m.¢m *mcH»muumq mm.mH m~.¢m *»cmcmm»m m@.Q~ mw.~> +Hm ~¢.H~ o>.ww Qm-m~ mm.w~ wH.>w ¢~-mH >H.w~ mm.m> @H-mH MUHMEQM m¢.m~ m~.~m +fim mm.¢~ H>.~> Qm-m~ mw.¢m Q@.m> ¢~-mH wm.m~ ~¢.>> \ mH-mH mwfimz mmm mmmm gum m@HnmH»m> wwmmzm aoom wm<\xwm .A©munmwm3v mQmwE msoum mmm Ucm xmm » 40¢ we wmmucwonwm m mm mmxmucfl UGUHHUDZ .m.m wfinme 36 Hm.om Q~.¢w am.wo ma.mm ~m.~> oo.>> om.oN oo.Hm m>.H@ wm.mm A a mH om.mm Nm.om mo.>m mH.om mH.>m mH.>m NH.m¢ oN.wm mmmm ZHU4HZ mm.Hm Hm.wm om.wNa ¢~.>¢H ¢¢.¢QH mh.ooa ¢m.~¢H @m.mHH w¢.H~H mm.wHH AQ4 .msucoE m umnfiw osu _cmEo: mcwumuoma now ~umum0EHuu Unfisu msu mnu .cmEo3 ucmcmmum now “mm5Hm> mcwumfiduamo now wumz hn Uwms mum mcwsoaflou mnh umuozi mm.mm HN.oN Hm.>m Hm.¢a wm.woa Nw.mN mm.ww mm.wN m¢.Hw ¢m.Hm Hm.aoH Nm.mm Nm.wHH HN.om m¢.mm ¢¢.m~ mm.>@ ¢>.o¢ w¢.m¢H ¢w.>~ mmm mm mw.>oH ¢>.m> ¢a.NNH ¢o.No Ho.woa Hm.m> mm.¢oH >>.~> oH.omH ww.w@ NH.mmH o¢.mo wm.oHa m¢.m> N¢.oNH ~>.m> mw.¢mH >~.¢m Add H H om.ma aH.¢a m>.¢~ mN.wN mw.am mm.om mo.om ¢¢.w~ mm.mm om.mm mmmm ZHZ.H@ mm.mm mm.mHH mH.mm mN.mm wm.NHH mm.>~H @H.HoH mm.>oH wm.-H AA4 om.ooa m>.moH m>.>mH Nm.moa ~>.mm Nv.wHH mm.NmH mQ.@~H hm.moH m¢.m¢H mmm m¢.>H w@.mH ~m.w~ ~H.om mw.mm oH.om wo.mN mo.mm Hm.~m mm.~¢ mm¢m U ZHZ4HH> H¢.moH oh.wfiH m@.HmH H>.>~H @m.-H mo.o¢H ¢m.o>H ¢~.@mH o>.~mH mw.¢wH AA< ¢mnHum»umA auflmcmmum +Hm om|mm ¢~-mH @H-mH mmfimeom +am omvmm ¢~-mH wH-mH mwamz mwHnmHum> om<\xom .»¢ou ~.m wflnma 37 mm.Nw oa.mm ~m.¢w om.wm om.mm >o.Hm m¢.@> ~w.>> o>.Nm ¢m.mw mmm mm.mH mN.m ¢m.¢H ¢w.mH m¢.mH ma.HN >¢.Q~ mm.>~ mw.mN mm.H~ mmmm 2DHUA4U mm.o> @¢.mw @m.~> w¢.~> mm.m¢ ¢>.>w m>.om Hm.mm mN.w> mo.Nm AA4 m@.¢wH ¢H.mm mm.mmH aH.wN mm.moN ¢m.mw ~o.H@H mm.mm m¢.HmH ¢>.ww wm.HmH ow.mm bm.mmN mm.mm -.m¢~ om.Nm wo.maN m>.wmH >m.¢m~ wN.mm mmm mmmm wflm ZHZ .mnucoE .nmE03 ucmcmwum now "mm:Hm> mm.mH~ om.aoN mm.o¢N mm.w- Ho.HoN HH.mmH mv.mNm >H.mom mm.Hmm mm.amm AA4 H~.>> Hm.h¢ mo.moa mo.¢w m¢.H> mN.mm ~H.mHH ¢H.mm om.Nm H¢.o~H mmm mH.wH ¢~.@ >h.NN mm.wN >a.>N H>.mN Hw.¢~ oN.am ~o.~¢ mN.wm mm2 uwms mum N>.mm >H.mm Nm.>Ha mm.Noa mm.am m>.>QH Hm.hNa om.mNH ¢m.w~H mv.omH Qqé Hm.mm mN.mm Qm.¢> mo.mm H>.Nm >m.mw ¢m.¢> wm.hm mw.>w om.mh oN.mH ¢@.oH ~m.wH m¢.mH mm.NN mm.HN >m.mH ow.mN mb.om ¢Q.¢~ m mw.>w om.mw ww.¢w mm.m> Nm.mm mN.Nm mm.ww wm.mm mm.Nm H¢.mm ‘i ~umumwEHnu Uufizu wnu mfisofiow 2E x58? Aqd ¢msflumuumA *ucmQmmum +Hm om|mN ¢~-mH mH|mH MUHMEUW +am om|mm ¢~-mH mH-mH mmfimz mwHnwHum> wm<\xmm .»nou ~.m MHQMH 38 w¢.m¢ mm.mm om.Hm w>.¢m ¢m.>¢ ¢¢.mm mm.vw mm.Hw ¢~.wm ~o.¢w mo.ma ~¢.QH ¢w.mH >¢.o~ No.mN >@.m~ wm.H~ >¢.m~ w~.¢m mm.mN mmmm UZHN mN.om mm.ow om.H> wm.mm Hm.ww Hm.~> Hm.>> ¢w.Hw om.¢w mm.hm AA4 om.¢> oh.mm ¢¢.mm >w.m¢ mm.w¢ wo.Nm ¢w.o~H NN.oHa N¢.Hoa mo.~HH w¢.wH mo.w w@.¢~ ~m.mH om.NN mm.mH m¢.~m mm.mm >w.>m ¢w.mm mmmm ZOMH .mnucoE w umufim msu .cmEo3 mcfiumuumfi uou Nnmumwewuu unfinu usu .soEo3 ucmnmwum now "mw5Hm> mcfiumfisofimu now mumz >n umms mum mcfisoflfiow oak “uuoz¢ ~H.¢w wm.>m mm.moa Nh.mm mm.mm w>.>w mm.o¢H ~m.¢¢H mm.m¢H HH.¢¢a AA4 >w.w¢ ~>.mm Q>.>w ¢m.>m m~.w¢ ~>.w¢ ¢m.mm mN.ow HN.Hm mN.mm mmm Nm.aH om.¢H ¢N.rH mN.mH mm.oN mm.mH >m.>H mm.aN mN.mm oa.HN mm >w.~> om.Hm mm.mw @>.w> w~.>> m¢.¢> ¢>.mw Add ~>.m@ mm.m> >m.oo >m.Hm NH.>m @>.>w wH.w~ mo.mN wm.m> ¢m.Hm m~.- mm.mH H ao.mN mN.mm Hm.w~ ¢o.mN A ¢~.@m a H¢.m¢ H~.w¢ m~.m¢ mmm mm.w> om.mm mN.m¢H Hw.~wH ¢m.wHH >m.>NH AA< amcwumuumq +ucmcmoum +Hm om|mm ¢N|mH manma mmameom +Hm ¢m-m~ ¢~-m~ oaumfi mmflmz mmanmHum> wm<\xuw .»cou ~.m MHQMH 39 mm.rw wH.o> mo.mw wH.~w wm.H> Nm.mm wm.mm HN.>w m¢.>w mH.mw m¢.mw oN.mw o¢.mm mm.>N mw.mN mm.>~ mm.mH Nm.HN NN.mN vH.bN m¢.m~ >H.w~ mm.¢N w¢.w~ w>.>~ mo.mN mmm mmmm M ZHZ@BH> om.mw om.mm Hm.>m wm.o> >m.~m ~>.ww NH.>m ~w.>w mN.wm ¢w.¢w w¢.mm ¢¢.mw -.mw mm.oHa >m.¢mH mm.¢HH ¢m.mm ~m.¢~H mm.>¢H oo.aNa m>.¢¢H mw.mHH >>.mHH mo.moH w~.¢H~ Hm.NNH AA< mmm mm.om wm.mm H>.om m>.¢m ¢¢.om Hm.om mm.om Hw.w~ >>.m~ No.mm Hm.mN ¢¢.om m>.~m mmmm 4 ZH24HH> m>.HmH ¢w.mmH m¢.wmH mH.>oH om.o¢H Nm.amH wm.H¢H Qm.¢~H wH.>mH wm.~¢H @w.m~H mN.wmH ¢H.m¢H AA4 am.ooH >m.mm Hm.mm mm.moa mm.woH mm.wm ¢m.HoH om.NoH ¢m.>m H>.mm ¢m.moa >m.mm Hm.wm mmm @¢.~¢ ~¢.~m H¢.H¢ >w.>m >m.¢m ¢m.m¢ mm.>m mm.mv Hm.Hv Nm.o¢ Na.a¢ ~¢.~¢ Hm.mm mmmm ZHMHOMm Hw.¢mH ~m.mHH m~.m~H m~.w~H ¢H.mNH Hv.HmH N>.>NH mw.~mH mm.>~H m¢.m~H mw.mNH No.omH ¢m.m~H Adé mm.wm m>.¢m >m.mm Hm.@m @¢.Hw Ho.mm mo.mm NH.>m ro.mm Nw.mm mm.wm m>.mm mo.¢m ¢w.¢~ m~.wH wH.¢~ mw.oN mm.oN >¢.m~ ¢w.- mm.mm >~.¢~ om.NN oo.¢N m~.¢~ mw.mN mmm mmmm MUMWZM QOOm ¢w.~> m¢.mm o~.m> m~.>w aN.N> >~.m> mm.H> N>.¢> mm.~> Hm.H> om.m> >¢.~> m~.o> OZ mmw umfia Hmfiuwmm unmH@Huwu|coz ucmfimwumm mEmum noom @m>oHmEm uoz vm>oHmEm mdumuw ucwE>0HmEm uwwz susow umwzufiz uwmosuuoz cofimmm ouuwE|coz cmnunndw >uHu Hmnucwu GOHUMNHCMQHD 0dQMHHM> .mm:onm qofiumflsmom umuomfiwm now | mam mo wmmucmonom m mm mwxmuqfl ucmfluusc mo mcmmz .m.m wfinme 40 mm.mm mm.oo ¢o.¢m Nm.mm NH.mo mN.mm m¢.Nm HH.wm >w.~m >m.¢m mm.wm Hw.mm wm.Nm H H mmm >m.mm Hm.mN H¢.mm mN.om w>.m~ Hm.bm wm.Nm NH.>m H¢.mm >~.vm HN.¢m mm.mm mm.¢m mmmm ZHUm.wHH >m.mHH wm.mHH No.wHH HA4 ¢m.mm wH.mm >H.wm oo.mm @¢.>o Nw.mm NN.mm Nm.mm oo.>m ¢¢.~m mm.Ho ¢¢.¢m Hm.mm Hw.om mm.NN mm.mN mm.mN H mm.mN NH.Hm w~.@~ wm.om mH.om o¢.mN H HH.om Hm.mN ¢w.m~ mmm mmmm zH>mqmomHm oH.>HH m>.mHH mN.>HH HN.woH No.HNH mH.¢HH w¢.wHH m>.wHH mw.mHH wv.NHH ww.oNH HH.wHH HH.mHH HA4 No.>m m¢.Hm >¢.>w Hm.mm mo.mm Hm.Hm mm.mm ¢o.om mN.mm om.mm m>.~m mm.wm bm.mm m¢.m~ mw.- w>.m~ mm.¢N m>.¢~ mo.om Hm.mN N¢.om mN.mN NH.mm ¢@.w~ mH.mN ~¢.>~ mmmm ZHZ¢HmH mw.hoH mw.moH Nm.boH ww.eoH wo.HHH mm.moH mH.NoH mm.oHH >m.moH w¢.moH hm.oHH wr.>oH mN.moH Add >w.cNH Nm.mmH wH.wNH Nm.voH mH.vmH Hm.mHH ¢w.>NH om.mHH mo.mHH ¢m.>mH wm.oHH o¢.omH Nm.mNH mmm w>.¢m >>.@~ wH.¢m mm.om ¢>.>~ m¢.wm mm.mm mm.Hm mm.om No.mm w~.~m mw.¢m w¢.¢~ mmmm U ZHZ4HH> mm.¢¢H wH.NhH NN.omH hw.hHH Nm.m¢H m¢.m¢H oo.¢mH HN.H¢H mN.H¢H mm.mwH mm.omH Hm.mmH Nm.NmH Add OZ mmw pawn Hmfiuwmm ucwHmH0uu|coz ucmHmHuom msmuw uoom ©o>oHmEm uoz ©m>0HmEm mnumum unmE>oHmEm um»: nuaom ummzvfiz ummwnuuoz noHmmm onuoE»¢oz nmnunnnm xuwu Hmnucwu GOMUMNHGMQHD UHDMHHM> .»nouL~.m mHQmH 41 ¢~.mw mm.mw mm.mw Hm.mm mm.ww mw.¢w >>.m> oo.am wH.>w Nm.No mm.mw m>.¢w ¢o.mw mmm Hm.NN wm.mH wH.- HN.wa w@.wH mw.mN ¢N.aN mw.H~ mo.mN mm.H~ m>.H~ wm.NN ¢a.aN mm4m ZDHUA mm.Hm mm.Nw m¢.m> mm.mw ¢~.m@ ~m.m> m>.mm m¢.>> ¢m.~m ¢o.Hw w¢.>> QA4 om.mHN mo.mm No.Hma ~m.¢m ¢w.mH~ w@.¢w mo.¢mN oo.mw m¢.>- ¢¢.@w @¢.>Q~ >o.om ao.moN om.mw N¢.NHN QQ.>> H¢.moN wm.w> H>.wm~ No.mHH oH.mHN mo.¢> mm.~H~ wm.Hm mm.H- ww.m> m¢m mm ~m.H@~ >w.mH~ >w.m>~ Hm.mm~ H¢.mw~ wm.m>N mh.omN wm.mmN mH.mmN ~¢.wHm mN.owN ~w.m>~ wm.m>~ Add Nm.¢m Nm.¢o NN.mm Hm.mm om.mo Qm.Hm m>.mm Nm.Ho mN.om m>.mm m¢.mm mm.mm Hm.mm H H a mmm mN.mN ~¢.- Nm.mN mm.NN m>.- >m.om >>.@~ om.mN H>.w~ om.wN wH.>N m>.w~ mm.wN mmmm memqom mm.¢HH hH.maH N¢.waH mm.mm m>.mHH H¢.maa N¢.mHH m>.fiNa mm.moa m>.NaH wm.~Hfi mm.bHH m~.mHH QA4 w¢.>m m>.Hh -.mw wm.¢m ¢m.m> @m.¢m ¢H.ww wH.mw mH.>w H>.mm m>.ww mH.mw m>.mw NN.NN Nm.>H ¢>.aN mm.mH mm.mH Hm.NN NH.HN N>.NN >m.H~ oH.oN mN.HN om.HN Hm.H~ mmm mm mo.mm mm.mm ow.mm mm.N> mH.mm >m.~m @m.~@ mn.mm mm.Nm >m.~@ aN.om ¢o.mw H¢.Hm Add oz mm» Umflo Hmflowmw uCwH@HUwH|GOZ ucwfimfluwm mEmum Uoom Uw>OHQEm uoz U0>0HQEm mfiumum uCUE>OHQEM umwz nusom uwwswfiz ummmnunoz cofimwm 0uumE»coz QMQHDQDW >uwu fimuucmu GOMUMNHGMDHD UHQMfiHM> .»cou m.m wflnme 42 bH.mm aw.wm m>.mm m>.Qw mm.Nm w¢.wm mh.mm mw.om om.>m mv.mm >~.Hw >m.>m ho.wm mH.¢~ >m.mH mm.mN Hm.- oN.oN m>.¢~ mH.H~ mw.mN >w.m~ NN.NN ¢¢.m~ mm.mN om.mN mm m¢.mw m¢.m> >>.o> mm.m> ¢H.w> m~.¢> w¢.@> ~w.¢> w@.H> >~.m> HN.mh Ho.¢h QA4 mo.mw mm.wm ~>.w@ mm.m> mw.Hm mN.mm m>.mw Hm.om ¢m.Hw Hm.¢@ mo.om mN.wm ~m.Hm mm.om mN.mN mm.mN mm.mN m¢.m~ wH.Nm ~@.m~ mm.om mm.mN mm.om w¢.@~ mm.om ~H.m~ mmofl >H.m¢H mm.>¢H mo.om NH.¢oa wm.moH >m.woH ¢>.aHa oN.NoH >¢.moH HH.woa mw.woa Ha.Hoa AA4 mm.wm -.¢w Nh.mm om.wm m¢.mw ~m.>m mm.Nm Nm.mm mo.mm mo.mm am.mm HH.ow ma.wm mmm >>.¢~ mm.wH ov.oN mm.ma mm.mH mw.H~ mv.oN mm.om >~.o~ wo.mH mm.mH >¢.oN w¢.Q~ mmmm 2DHmWZU42 ¢¢.m> ¢m.¢h >a.¢> H~.mw mH.Nh N¢.¢h mm.w> om.m> mm.Nh w¢.~> mm.H> mm.¢> mQ.~> Qqfi oa.Noa N>.wm mh.NoH w~.m~ ¢m.~oH >w.>m mm.mm @¢.m~ oh.moa wm.@~ Nm.HoH ~¢.H¢ mH.moH HN.mm mw.ooH ~H.mm H~.H¢H Nw.mm oN.mm >H.mm mo.woH Nm.mm Nm.aoH ¢o.mm Hw.mm >fi.>m mmm mmmm wnmommwomm mm.mNa wm.HNH m¢.m~H >w.@oH >m.mHH mm.mma m~.~m~ m>.>~H >m.w~H Hm.m~H mm.mNH No.mNa No.mNH AQ4 OZ mow uwfln Hmfluwmw uc0wmHumn|coZ ucwwmfiumm memum uoom Uo>oHmEm uoz ©0>OH&Em mdumuw uGUE>OH@Em umwz susow ummauflz ummosunoz nowmmm 0HUUE|GOZ cmnnsnsm >ufiu Hmuucmu COHUMNHGMQHD UHDMHHM> .uaou ~.m MHAMH 43 APPENDIX C. Parameter Estimates of the Nutrient Demand Models 44 p**;\ Variable Intercept Height Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq Weekend Income lncomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type Note: Numbers rounded to three decimal places. FOOD ENERGY ALL FAFH (KILOCALORIES) FAH 869.750* 965.389* 184.184 (179.410)(161.272)(191 0.790* (0.260) 14.680* (2.760) -36.850* (21.150) -21.020 (17.780) -2.050 (20.980) -37.530* (19.590) -58.130* (21.140) -60.760* (27.160) -15.860 (80.640) 116.550* (55.880) -49.730 (46.860) 480.770* (21.370) -2.230 .330) .170 .460) .080* .180) -5.660 (5.640) .310* (2.190) 0.090* (0.020) .320* .170) a (0.001) - a ( a ) 0.2400 0.2000 0.0160 0.0190 uls -8. <18. -14. <15 -18. <18. -41 <18. -82. <18. 23. <24. 0 <83. 124. <81 -58. <48. 154. <18. 81 <14. -81 <35 -79. <17. -38. <4. -22. (1 _ 0. (0. 10. .512) (21 001* -0. 001) (0. (16 0. (0. ( 0.1294 0.6414 0.0038 0.1823 .914* 0. .234) <0. .138 13. .500) <2. -43. 898) <22. -13. .214) <18. 30. 006 322 054 309) (21 .442* -9 547) (20. 337* 1. 313) (21 583 786) (81 149* 102. .492) (57. 27. 595) (47. 922* 375. 815 775) (22 .676*-129 790) (17 .426* 24. .530) (36. 205* -88. 927) (20. 38. 091* 832) (5 747* 6. .996) (2. 192* -0. 021) (0. 891 -11 a * 8 ) ( uls -22. 848) <28. .878 24. .800) 125 271) 529* 973) 424* 041) 468 690) 773 .601) .455 362) 526 .947) 913 003) 258 .296) 743* 893) 524 152) 820* .584) .411* .958) 397 757) 335* 623) 863* .776) 065* 220) 052* 022) .049 .046) 003* 001) a i a ) 0.1322 0.1346 0.3355 0.2559 uls ALL 13. <0. 0. <0. -0. <0. 0. <0. 0. <0. 0. <0. 0. <0. 0. <0. 0. <0. 1 <0. 0. <0. 0. <0. 0. <0. -0. <0. (0. PROTEIN FAFH (GRAMS/1000 KILOCALORIES) 505* 963) 005* 001) 044* 014) 186* 113) 024 095) 138 115) 314* 104) 055 115) 411* 148) .062* 390) 032 269) 899* 220) 042 112) 024 094) .210 199) .050* .126) .024 .029) .098* .011) .001* a ) .219* .107) a a ) a a ) 0.0459 0.2117 0.0206 0.0029 uls 13. <2. 0. <0. -0. 934* 156) 003 003) 051 .033) .201 .248) .226 .210) .067 .253) .522* .223) .111 .251) .098 .338) .183 .020) .775 .669) .230 .644) .195 .246) .248 .213) .036 .517) .031* .294) .244* .066) .081* .026) .001* a ) .303 .222) ‘k a a ) a * a ) 0.0267 .1 DDD 636 .0529 .6858 wls Numbers which are "Q" indicate values less than 0.0005. Figures in parentheses are standard errors. Imratio denotes the inverse of Mill's ratio. wls denotes weighted least squares. ols denotes ordinary least squares. Asterisk denotes statistical significance at the 0.05 level. 45 FAH 13. (1 0. (0. -0. (0. 0. (0. 0. (0. ( 0.0 0.1 0.3 0.1 643* .154) 006* 002) 045* 017) 288* 138) 071 114) .072 .139) .216* .125) .050 .140) .328* .178) .640 .470) .236 .315) .168* .244) .154 .134) .018 .113) .202 .227) .095* .146) .008 .034) .093* .013) .001* a ) .386* .129) a a ) a a ) 313 012 692 223 uls ALL 13. (0. (0. 0.0 DDD FAT 239* 757) .001 001) .007 .011) .568* .091) .211* .075) .017 .091) .293* .081) .042 .092) .171 .115) .075* .377) .032 .209) .312* .180) .181* .088) .183* .074) .048 .154) .762* .098) .041* .023) .036* .009) a *- 8 ) .036 .085) a 6 ) a 8 ) 339 .0001 .0004 .0190 uls FAFH (GRAMS/1000 KILOCALORIES) 14.720* (1.725) 0.003 <0.002> -0.015 <0.027> -0.871* <0.202> -0.585* <0.188> -0.098 <0.203> 0.388* <0.180> -0.128 <0.208> -0.550* <0.282> 0.281 <0.836> -0.321 <0.537> -0.278 (0.475) -0.711* <0.198> -0.089 <0.171> -1.028* <0.397> 0.158 (0.223) -0.069 (0.054) -0.005 (0.021) a 8 ) 0.140 (0.173) DDD a * 8 ) a ‘k 8 ) .0220 .0001 .0398 .1807 uls FAH 13 <0. 0. <0. -0. <0. -0. <0. -0. <0. 0. <0. 0. <0. 0. <0. -0. <0. -1 .268* 915) 001) 014 014) 564* 109) 181* 091) 086 109) 438* 098) 027 112) 012 139) .118* .442) .317 .257) .197 .222) .404* .106) .082 .089) .032 .183) .802* .114) .113* .028) 0.034* 0.0 .0 .0 .0 DDD .011) a * .029 .103) mill!) 323 001 001 519 uls Variable SAT. FAT ACIDS MONO FAT ACIDS POLY FAT ACIDS ALL FAFH FAH ALL FAFH FAH ALL FAFH FAH (GRAMS/1000 KILOCALORIES) (GRAMS/1000 KILOCALORIES) (GRAMS/1000 KILOCALORIES) Intercept 5.230* 5.286* 5.361* 5.150* 5.629* 5.123* 1.849* 2.460* 2.002* (0.348) (0.742) (0.413) (0.317) (0.680) (0.386) (0.273) (0.633) (0.310) Height 0.001 0.001 a 0.001* 0.002* 0.001 -0.001* - a -0.001 (0.001) (0.001) (0.001) ( a ) (0.001) (0.001) ( a ) (0.001) ( a ) Height -0.001 0.001 -0.006 -0.007 -0.006 -0.010* 0.003 -0.003 - a (0.005) (0.011) (0.006) (0.005) (0.010) (0.006) (0.004) (0.010) (0.005) Urban1 -0.196* -0.294* -0.200* -0.281* -0.306* -0.284* -0.047 -0.150* -0.044 (0.041) (0.086) (0.050) (0.037) (0.080) (0.045) (0.033) (0.076) (0.036) Urban2 -0.118* -0.214* -0.144* -0.111* -0.204* -0.099* 0.036 -0.131* 0.073* (0.035) (0.072) (0.042) (0.031) (0.066) (0.038) (0.027) (0.065) (0.031) Region1 0.215* 0.345* 0.206* -0.139* -0.190* -0.083* -0.088* -0.206* -0.036 (0.042) (0.088) (0.050) (0.038) (0.079) (0.046) (0.033) (0.075) (0.037) Region2 0.251* 0.343* 0.274* 0.055 0.110 0.106* -0.031 -0.127* 0.023 (0.037) (0.077) (0.045) (0.033) (0.071) (0.041) (0.029) (0.067) (0.034) Region4 0.110* 0.096 0.134* -0.118* -0.113 -0.091* -0.004 -0.065 0.015 (0.042) (0.087) (0.051) (0.039) (0.081) (0.047) (0.034) (0.077) (0.038) Race2 -0.218* -0.274* -0.185* 0.091* -0.123 0.179* -0.084* -0.110 -0.069 (0.052) (0.110) (0.062) (0.048) (0.106) (0.058) (0.040) (0.092) (0.045) Race3 -0.859* 0.117 -0.911* -0.421* -0.124 -0.407* 0.297* 0.323 0.304* (0.155) (0.386) (0.170) (0.156) (0.306) (0.181) (0.141) (0.338) (0.161) Race4 -0.035 -0.322 -0.100 -0.005 -0.238 -0.071 0.011 0.136 -0.101 (0.097) (0.223) (0.116) (0.086) (0.210) (0.109) (0.077) (0.197) (0.085) Hisp1 -0.241* -0.234 -0.176* -0.008 0.001 0.015 -0.157* 0.027 -0.154* (0.082) (0.197) (0.101) (0.074) (0.186) (0.093) (0.063) (0.175) (0.071) Sex1 -0.005 -0.213* 0.068 0.203* -0.192* 0.290* -0.042 -0.322* 0.032 (0.040) (0.085) (0.048) (0.036) (0.078) (0.044) (0.032) (0.072) (0.036) Employ1 0.039 -0.070 0.018 0.075* -0.022 0.051 0.067* 0.093 0.029 (0.034) (0.074) (0.041) (0.031) (0.067) (0.037) (0.027) (0.063) (0.031) Fstamp1 -0.061 -0.314* -0.051 0.075 I-0.254 0.112 -0.058 -0.253* -0.033 4 (0.069) (0.167) (0.081) (0.064) (0.163) (0.076) (0.054) (0.125) (0.060) diet1 -0.427* -0.174* -0.437* -0.317* -0.026 -0.307* 0.014 0.290* -0.009 (0.043) (0.094) (0.050) (0.041) (0.087) (0.048) (0.035) (0.091) (0.038) Hsize 0.011 0.016 0.038* 0.037* -0.007 0.061* -0.013* -0.070* 0.005 (0.010) (0.023) (0.012) (0.010) (0.021) (0.011) (0.008) (0.019) (0.009) Age -0.012* -0.025* -0.011* 0.012* -0.011 0.014* 0.030* 0.022* 0.027* (0.004) (0.009) (0.005) (0.004) (0.008) (0.005) (0.003) (0.008) (0.004) Agesq a a * a - a * 0.002* - a * - a * - a - a * ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) weekend -0.012 0.038 -0.018 0.048 0.067 0.046 -0.007 -0.040 -0.002 (0.039) (0.076) (0.047) (0.035) (0.068) (0.043) (0.031) (0.063) (0.034) Income a a * a - a a * - a * a * a * - a (a)(a)(a)(a)(a)(a)(a>(a)(a) lncomesq - a - a * - a * - a - a * - a - a - a * - a ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( 8 ) ( a ) ( a ) Imratio* Adj R-SQ 0.0562 0.0200 0.0459 0.0474 0.0157 0.0428 0.0295 0.0464 0.0130 F-Tests Prob >F Urbanization 0.0001 0.0012 0.0001 0.0001 0.0003 0.0001 0.0252 0.0761 0.0019 Region 0.0001 0.0001 0.0001 0.0001 0.0019 0.0001 0.0450 0.0390 0.4860 Race 0.0001 0.0434 0.0001 0.0101 0.4580 0.0014 0.0266 0.3956 0.0560 Regression Type wls wls wls wls wls wls wls wls wls 46 -’”\ Variable Intercept Height Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq weekend Income Incomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type CHOLESTEROL ALL FAFH (MILLIGRAMS) .254* .360) 0.347* .069) 1.879* .749) .170* .650) .237* .680) .930 .413) .138* .088) .003* .553) .746* .674) .782* .254) .774* .659) .265* .883) .227* .670) .092 .490) 2.700 .782) .951* .338) 1.877 .499) 0.459 .563) .007 .006) .860* .268) .001* ( a ) a * ( a ) 0.1387 0.0209 0.0069 0.0001 wls 82. (34. 0. (0. 0. (0. 2. (4. -0. (3. 1 (4. -4. (3. -7. (4. 5. (5 19. (16. 21 (13. -2. (11 20. (4. 12. (3. -1 (7. -12. (3. -6. <1 -1 (0. 0. (0. 5 (3. ( ( 0. 738* 796) 193* 050) 234 543) 817 154) 448 338) .802 138) 542 521) 136* 121) 097 .591) 156 309) .453 108) 561 .612) 750* 073) 116* 290) .235 848) 127* 995) 396* .047) .831* 439) 016* 005) .609 673) a 6 ) a 8 ) 0495 0.6825 0. 0. 1482 2047 wls CARBOHYDRATE DIETARY FIBER FAH ALL FAFH FAH ALL FAFH FAH (GRAMS/1000 KILOCALORIES) (GRAMS/1000 KILOCALORIES) 22.690 42.196* 40.639* 43.505* 1.649* 2.723* 1.682* (47.864) (2.333) (4.975) (2.844) (0.292) (0.507) (0.360) 0.188* -0.010* -0.013* -0.009* -0.001* 0.001 -0.001* (0.068) (0.003) (0.007) (0.004) ( a ) (0.001) (0.001) 1.609* 0.045 0.035 0.049 0.005 -0.011 0.007 (0.741) (0.036) (0.077) (0.044) (0.005) (0.007) (0.006) -17.154* 0.427 0.556 0.829* 0.088* -0.073 0.101* (5.594) (0.272) (0.594) (0.332) (0.035) (0.059) (0.045) -11.154* 0.145 0.106 -0.011 0.065* -0.081 0.084* (4.664) (0.226) (0.477) (0.277) (0.029) (0.049) (0.036) 2.265 -0.414 -0.608 -0.134 -0.089* -0.151* -0.086* (5.264) (0.261) (0.591) (0.312) (0.035) (0.058) (0.043) -10.051* -1.057* -2.146* -1.127* -0.016 -0.031 -0.027 (5.069) (0.245) (0.503) (0.301) (0.031) (0.052) (0.040) -2.264 -0.162 0.117 -0.522 0.074* 0.108* 0.037 (5.522) (0.268) (0.589) (0.328) (0.037) (0.059) (0.046) 47.306* -0.207 2.101* -0.873* -0.257* -0.063 -0.326* (7.771) (0.370) (0.799) (0.461) (0.043) (0.078) (0.053) 39.679* 0.932 0.401 0.797 -0.037 -0.087 -0.119 (17.91?) (0.881) (2.332) (1.064) (0.158) (0.228) (0.182) 43.304* 0.755 1.313 1.538 0.161* -0.066 0.209* (16.183) (0.755) (1.874) (0.961) (0.082) (0.160) (0.101) 55.844* -0.148 0.561 0.083 0.333* 0.067 0.298* (13.139) (0.621) (1.660) (0.780) (0.077) (0.142) (0.088) 70.128* -1.337* 0.873 -1.799* -0.086* -0.083 -0.079* (5.654) (0.274) (0.582) (0.336) (0.034) (0.057) (0.042) -23.122* -0.480* 0.686 -0.226 -0.101* -0.087* -0.031 (4.415) (0.216) (0.470) (0.262) (0.029) (0.050) (0.035) 6.934 -0.536 1.830 -0.633 -0.096* -0.062 -0.110 (9.683) (0.472) (1.122) (0.575) (0.055) (0.115) (0.073) -19.457* 1.437* -1.256* 1.687* 0.413* 0.259* 0.449* (5.041) (0.257) (0.571) (0.299) (0.040) (0.062) (0.048) 9.221* 0.098 0.682* -0.219* -0.038* -0.039* -0.056* (1.491) (0.072) (0.149) (0.088) (0.009) (0.015) (0.011) 2.070* -0.215* -0.141* -0.218* 0.021* -0.008 0.020* (0.561) (0.027) (0.062) (0.033) (0.004) (0.006) (0.005) -0.017* 0.002* 0.001 0.002* - a * a * - a (0.006) ( a ) (0.001) ( a ) ( a ) ( a ) ( a ) 10.348* -0.252 -0.777 0.264 -0.080* 0.040 -0.100* (5.170) (0.254) (0.525) (0.307) (0.032) (0.052) (0.042) -0.001* - a - a * a a * a * a * ( a ) ( a ) ( a ) ( a ) ( a ) ( a ) ( 8 ) a * a a a - a - a * - a ( a ) ( 8 ) ( 8 ) ( a ) ( 8 ) ( a ) ( 8 ) 0.1107 0.0387 0.0352 0.0370 0.1375 ‘0.0205 0.1001 0.0059 0.2880 0.6184 0.0130 0.0222 0.2390 0.0292 0.1104 0.0002 0.0001 0.0013 0.0010 0.0015 0.0664 0.0001 0.4833 0.0631 0.0698 0.0001 0.8252 0.0001 uls uls uls uls uls ols uls 47 Variable ALCOHOL VIT A-IU CAROTENES u ALL FAFH FAH ALL FAFH FAH ALL FAFH FAH (GRAMS/1000 KILOCALORIES) (INTER. UNITS) (MICROGRAMS RE) Intercept 2.619 3.025 5.561 885.590 269.050 334.658 -37.705 80.167‘ -73.299 (2.404) (9.514) (3.921) (1771.7)(943.165) (1700.8)(151.630) (74.530)(145.572) Height 0.001 -0.013 0.006 -1.059 1.034 -1.083 -0.194 0.136 -0.216 (0.003) (0.013) (0.006) (2.564) (1.320) (2.468) (0.219) (0.105) (0.211) Height -0.011 0.035 -0.028 48.499* 27.904* 34.633 3.893* 0.765 2.863 (0.036) (0.139) (0.058) (27.462) (14.731) (26.360) (2.350) (1.154) (2.256) Urban1 0.804* 2.813* 0.233 475.170* 179.598 389.096* 26.574 4.412 22.102 (0.276) (1.045) (0.454)(212.000)(115.764)(203.649) (18.144) (9.116) (17.429) Urban2 0.699* 1.369 0.770* 228.940 -132.115 341.503* 21.209 -10.494 30.114* (0.250) (0.987) (0.404)(177.571) (93.333)(170.551) (15.197) (7.292) (14.596) Region1 -0.437* 1.739* -0.979* 656.758* 114.429 749.705* 50.978* -3.293 57.825* (0.263) (1.027) (0.416)(210.463)(112.926)(201.859) (18.012) (8.857) (17.276) Region2 -0.266 1.738* -1.051* 73.649 -270.525* 261.787 -5.388 -27.517* 11.757 (0.266) (1.009) (0.431)(192.921) (99.618)(185.063) (16.511) (7.746) (15.838) Region4 -0.274 1.679* -1.013* 437.205* -46.725 543.980* 54.241* -1.155 60.044* (0.262) (1.017) (0.416)(214.907)(113.015)(206.533) (18.393) (9.185) (17.676) Race2 0.201 -2.958 0.999 -571.997*-389.547*-287.668 -59.504* -19.998* -41.246* (0.446) (1.856) (0.683)(267.118)(141.432)(256.735) (22.861) (11.186) (21.972) Race3 0.108 2.915 -0.732 556.829 425.729 399.769 151.470* 65.659* 128.758* (1.261) (5.276) (1.909)(790.672)(411.208)(757.116) (67.671) (33.123) (64.798) Race4 -0.849 2.424 -0.697 979.115* 517.823* 929.043* 108.340* 35.836 99.959* (0.881) (3.028) (1.598)(511.856)(290.052)(497.688) (43.808) (25.034) (42.594) Hisp1 -0.526 -0.785 -1.204 -419.029 -69.891 -274.189 -12.330 -1.841 -4.580 (0.565) (2.111) (0.884)(435.681)(260.579)(419.710) (37.288) (22.642) (35.921) Sex1 0.893* 2.697* 0.724 717.288*-111.460 717.772* 21.149 -2.678 23.637 (0.272) (1.004) (0.457)(206.828)(106.896)(198.563) (17.701) (8.486) (16.994) Employ1 -0.175 -2.686* 0.298 -239.349 391.630*-610.425* -14.838 23.689* —38.751* (0.240) (0.892) (0.393)(174.706)(101.189)(167.631) (14.952) (7.228) (14.346) Fstamp1 1.499* 2.582 2.021*-212.036 »5.942 -243.918 -65.966* 2.338 -63.392* (0.726) (2.626) (1.148)(349.720)(222.481)(335.907) (29.931) (16.713) (28.748) diet1 -0.178 0.616 -0.325 747.774* 42.106 855.180* 92.316* 8.316 92.113* (0.283) (1.037) (0.470)(214.151)(119.599)(205.623) (18.328) (9.201) (17.598) Hsize -0.148* 0.007 -0.210 -199.595* -83.960* -76.771 -13.726* -7.067* -4.785 (0.079) (0.337) (0.129) (56.575) (28.31?) (54.412) (4.840) (2.248) (4.656) Age 0.012 0.067 -0.007 30.415 -64.986* 61.935* 4.830* -3.931* 7.016* (0.035) (0.153) (0.061) (22.086) (13.778) (21.275) (1.890) (1.027) (1.820) Agesq -a -0.001 -a 0.071 0.850* -0.277 -0.019 0.054* -0.043* ( a ) (0.002) ( a ) (0.233) (0.159) (0.224) (0.019) (0.011) (0.019) weekend 0.093 -0.073 -0.544 -677.648* -76.914 -770.372* -60.082* -2.417 -67.841* (0.228) (0.786) (0.370)(197.634) (92.966)(189.513) (16.914) (7.625) (16.219) Income - a - a * - a 0.020* 0.008* 0.007 0.002* 0.001* 0.001* ( a ) ( a ) ( a ) (0.006) (0.004) (0.006) (0.001) ( a ) (0.001) Incomesq a a a - a - a - a - a * - a - a (a)(a)(a)(a)(a)(a)(a)(a>(a) Imratio* 0.773* 2.194* 0.149 (0.319) (0.835) (0.489) Adj R-SQ 0.0262 0.0845 0.0303 0.0402 0.0207 0.0383 0.0412 0.0209 0.0377 F-Tests Prob >F Urbanization 0.0068 0.0252 0.1074 0.0798 0.0104 0.0781 0.2558 0.1313 0.1150 Region 0.4189 0.2485 0.0310 0.0068 0.0034 0.0014 0.0007 0.0013 0.0004 Race 0.7577 0.2864 0.4604 0.0260 0.0060 0.1537 0.0004 0.0215 0.0039 Regression Type heckman heckman heckman ols wls ols ols wls ols 48 ‘\.A Variable Intercept Height Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq weekend Income Incomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type VIT E-ALPHA ALL FAFH FAH (MILLIGRAMS ATE) 4.039* 2.039* 2.848* (1.695) (0.853) (1.690) a 0.002* -0.003 (0.002) (0.001) (0.002) 0.050* 0.016 0.033 (0.026) (0.013) (0.025) 0.260 -0.026 0.207 (0.195) (0.099) (0.197) 0.106 -0.069 0.143 (0.166) (0.082) (0.165) -0.064 -0.331* 0.272 (0.198) (0.096) (0.198) -0.344* -0.352* -0.045 (0.182) (0.088) (0.182) -0.030 -0.211* 0.193 (0.201) (0.108) (0.197) -1.345* -0.031 -1.061* (0.218) (0.130) (0.216) -0.011 0.474 -0.043 (0.670) (0.423) (0.622) 0.523 0.519 0.487 (0.513) (0.323) (0.520) -0.377 -0.249 -0.037 (0.425) (0.286) (0.424) 1.945* 0.304* 1.831* (0.192) (0.098) (0.197) 0.113 0.367* -0.573* (0.157) (0.080) (0.159) 0.060 -0.324* -0.016 (0.278) (0.166) (0.312) 0.004 -0.211* 0.125 (0.196) (0.101) (0.194) -0.172* -0.147* -0.013 (0.051) (0.025) (0.050) -0.021 -0.051* 0.033 (0.021) (0.010) (0 021) a a * - a ( a ) ( a ) ( a ) 0.033 -0.018 -0 197 (0.180) (0.089) (0 180) a * a * a ( 8 ) ( 8 ) ( a ) - a i - a - a i ( a ) ( a ) ( 8 ) 0.0603 0.0474 0.0417 0.4121 0.6887 0.5373 0.2545 0.0004 0.3605 0.0001 0.2759 0.0001 wls wls uls VIT C ALL FAFH (MILLIGRAMS) 23.796 24.676* (19.352) (9.504) 0.027 0.019 (0.027) (0.013) 0.699* 0.012 (0.299) (0.146) 4.402* 0.245 (2.249) (1.092) 6.761* -0.188 (1.877) (0.881) 14.304* 1.095 (2.331) (1.124) 1.564 -1.418 (2.017) (0.893) 5.229* 2.056* (2.371) (1.196) 20.679* 7.164* (3.062) (1.771) 20.342* 3.881 (9.020) (4.141) 30.844* 2.734 (6.245) (3.513) 4.390 -1.188 (4.973) (2.838) 10.555* 2.856* (2.257) (1.089) -8.289* 1.718* (1.887) (0.889) -9.481* -0.212 (3.417) (2.178) 15.031* -0.283 (2.423) (1.098) -3.249* -1.385* (0.591) (0.291) -0.297 -0.431* (0.239) (0.121) 0.005* 0.004* (0.003) (0.001) -8.010* 0.195 (2.058) (0 962) 0.001* a * ( 8 ) ( 8 ) - a "k _ a ( 8 ) ( 8 ) 0.0697 0.0228 0.0014 0.9105 0.0001 0.0166 0.0001 0.0007 uls uls 49 FAH .712 .901) 0.014 .026) 0.522* .292) 4.190* .187) 6.534* .840) .931* .276) 2.821 .982) 4.381* .324) 17.422* (2.942) 19.042* (8.813) 30.686* (6.240) 6.248 (4.879) 9.090* (2.209) -11.972* .844) -5.860* (3.374) 15.292* (2.385) -1.728* (0.577) (0.234) 0.001 (0.002) -10.028* (1.978) | IDDJO) Viv 0.0678 0.0018 0.0001 0.0001 wls ALL COO a 0.002 THIAMIN FAFH (MILLIGRAMS) 0.764* .176) 0.001* a ) 0.007* .003) .027 .020) .016 .017) 0.004 .020) .021 .019) .095* .020) .059* .025) 0.072 .074) 0.131* .053) .022 .043) 0.377* .020) .024 .017) .020 .035) .047* .020) 0.002 .006) .009* .002) a * ) .019) a a ) a a ) .1364 .3911 .0001 .0044 uls COO 0.626* .111) 0.001* a ) .001 .002) .012 .013) .012 .011) .012 .013) .038* .011) .042* .012) 0.031* .018) .006 .040) 0.053 .044) .015 .033) 0.100* .012) 0.058* .010) .024 .025) .048* .012) .023* .003) .013* .001) a 9: a ) 0.008 .011) a vivi- a 8 8 .1016 .4680 .0010 .2374 uls FAH 0.403* .180) ( a ) 0.007* .003) .022 .021) .011 .017) 0.025 .021) 0.001 .019) .055* .021) .048* .026) 0.094 .073) 0.130* .054) .044) 0.316* .021) .104* .017) 0.0949 0.5880 0.0063 0.0103 wls Variable Intercept Height Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq Heekend Income Incomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type ALL DDD RIBOFLAVIN FAFH (MILLIGRAMS) 1.018* .237) .001* a ) .013* .004) .014 .028) .029 .023) .049* .027) .032 .025) .011 .028) .168* .035) .033* .094) .093 .069) .134* .056) .425* .027) .060* .022) .022 .049) .043 .027) .011 .007) .020* .002) a i 8 ) .006 .025) a a ) a a ) .1195 .1835 .2803 .0001 uls 0.701* .142) .001* a ) a .002) .005 .016) .008 .013) .003 .016) .018 .014) .033* .016) .007 .020) .018 .058) .055 .058) .004 .039) .105* .016) .072* .013) .010 .030) .059* .015) .030* .004) .018* .002) a i a ) .002 .014) a i a ) a i a ) 0.0888 .6895 .1224 .7590 DDD wls FAH 0. (0. ( 0. (0. 0. .029) (0 -0. (0. 0. (0. 0. (0. 0. .028) (0 -0. .036) (0 -0. (0. 0. (0. -0. (0. 0. (0. -0. (0. 0. (0. 0. (0. 0. (0. -0. (0. ( 643* 242) a 8 ) 010* 004) 006 023 023) 058* 027) 048* 025) 044 125* 277* 100) 089 071) 078 057) 363* 028) 161* 023) 003 050) 005 028) 018* 007) 005 003) a i 8 ) .022 .026) a i ) a i a ) 0.0821 0.4653 0.1118 0.0001 wls ALL 0. DDD 50 NIACIN FAFH (MILLIGRAMS) .492* .300) .010* .003) .100* .035) .211 .272) .158 .225) .253 .276) .263 .247) .250* .272) .585* .345) .683 .844) .727* .676) .733 .570) .549* .272) .109 .221) .107 .462) ' .404 .265) .189* .071) .024 .028) a a .007 .257) a ) 8 ) a a ) 1642 .6901 0001 0149 wls 6. (1 0. (0. 0. (0 -0. (0. -0. (0. -0. (0. -0. (0. -0. (0. 0. (0. (0. 0. DDD 471 .722 009 002 013 .026 067 198 050 161 251 198 208 176 693 199 597 292 .245 712 .301 .638 .366 .525 .505 .201 i ) i ) ) ) ) ) ) i ) i ) ) i ) ) i ) .837* .160 .521 .348 .670 .189 .410 .050 .150 .021 .001 a .160 .175 a a a a 0984 9307 .0069 0443 wls ) ) i ) i ) i ) i ) ) i ) * ) .228* .390) .005 .004) .074* .037) .120 .285) .125 .236) .591* .287) .128 .258) .632* .281) .431 .357) .947 .854) 1.511* .721) .012 .589) .563* .285) .396* .229) .424 .475) .141 .277) .239* .073) .166* .028) .001* 8 ) .508* .268) DO) vI-w 0.1040 0.8542 0.0018 0.0615 wls V ALL IT B6 FAFH (MILLIGRAMS) .763* .216) .001* a ) .009* .003) .052* .025) .027 .021) .055* .025) .013 .023) .031 .025) .110* .030) .098 .073) .151* .061) .055 .051) .417* .025) .039* .020) .049 .040) .019 .026) .026* .006) .005* .003) a i 8 ) .025 .023) a i 8 ) - a a ) 0.1164 .1119 .0218 .0001 DDD wls .557* .125) .001* a ) a .002) a .014) .014 .011) .016 .014) .019 .012) .031* .014) .022 .020) .033 .051) .101* .046) .027 .036) .106* .014) .056* .011) .004 .027) .042* .013) .035* .003) .011* .002) a i 8 ) .023* .012) a i 8 ) a a ) 0.0889 .3701 .1679 .1048 DDD wls FAH .478* .222) a ) .007* .003) .043* .026) .027 .021) .082* .026) .021 .023) .002 .026) .093* .031) .046 .080) .126* .064) .001 .053) .355* .026) .124* .021) .041) .057* .026) .004 .007) .004 .003) 8 ) .068* .024) DOD) 0.0839 0.2220 0.0098 0.0033 wls W v3 Variable Intercept weight Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq Heekend Income Incomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type FOLATE ALL FAFH (MICROGRAMS) .110* .740) .034 .054) .769* .584) .760 .519) .080 .646) .074 .455) .210* .044) .755* .508) .651* .389) .579 .689) .258* .445) .297* .362) .131* .421) .425* .646) .174 .070) .579* .494) .036* .125) .004* .472) .014* .005) .809 .971) .001* a ) 0.0755 0.3736 0.0001 0.0014 uls 60. (18. 0. (0. 0. (0. 0. (2. -2. (1 -2. (2. -6. (1 -2. (2. 2. (3. 9. (7. 12. (7. -2. (5 0. COD 915* 447) 048* 025) 321 285) 133 212) 365 .784) 903 139) 966* .881) 957 215) 981 043) 982 195) 937* 214) 351 .450) .583* .080) .904* .725) .131 .383) .872* .009) .503* .556) .553* .245) .013* .003) .131 .816) a i a ) a i a ) 0632 2835 0033 1263 wls .1s2* .962) .011 .055) .34a* .569) .306 .545) .696 .691) .157 .457) .620* .064) .569* .567) .791 .377) .575 .019) .a31* .667) .25s* .571) .516* .466) .527* .669) .022 .017) .763* .590) .990 .111) .379 .472) .004 .005) .463* .011) 0.0679 0.4409 0.0011 0.0122 wls VIT B12 ALL FAFH (MICROGRAMS) .707 .585) .007* .004) .036 .040) .216 .296) .200 .246) .195* .329) .035 .259) .031 .291) .304 .420) .333 .726) .030 .800) .233 .625) .511* .301) .172 .242) .171* .660) .652* .302) .129 .084) .056* .031) .001* 8 ) .451 .294) a 8 ) a a ) 0.0172 0.3150 0.0014 0.2683 uls 51 2 (1. (0. 0. (0. 0. (0. 0. (0. (0. .527 724) a 002) 002 026) 076 202) 009 168) .355* 198) .085 .179) .018 .202) .293 .266) .175 .776) .021 .545) .072 .482) .592* .196) .436* .170) .059 .392) .617* .213) .147* .053) .070* .022) .001* 8 ) .098 .177) a a ) a 8 ) 0.0111 0.9176 0.0847 0.7337 ols CALCIUM FAH ALL FAFH FAH (MILLIGRAMS) 1.587 492.220* 319.868* 129.065 (2.315)(104.643) (66.637)(108.600) 0.006 0.069 0.192* -0.116 (0.004) (0.152) (0.091) (0.151) 0.006 7.027* 1.332 7.901* (0.035) (1.608) (1.012) (1.677) 0.295 10.611 4.260 11.530 (0.268) (12.973) (7.595) (12.773) -0.065 -8.414 -10.910* -4.400 (0.226) (10.761) (6.300) (10.605) 1.118* 39.800* 9.260 38.266* (0.296) (12.471) (7.549) (12.263) 0.317 47.925* -10.410 50.045* (0.237) (11.743) (6.632) (11.603) 0.195 76.976* -15.292* 88.454* (0.266) (13.46?) (7.812) (13.181) 0.571 -127.680* -12.370 -109.782* (0.410) (14.438) (9.423) (14.153) -1.266 -125.970* 20.884 -111.744* (1.055) (43.174) (28.586) (39.666) -1.142 -1.810 26.162 1.983 (0.769) (32.751) (29.720) (31.289) 0.277 -95.571* -15.355 -61.301* (0.574) (25.385) (21.415) (23.782) 1.232* 146.138* 40.023* 110.498* (0.265) (12.694) (7.785) (12.751) -0.172 -23.737* 31.853* -65.831* (0.218) (10.480) (6.169) (10.258) 1.361* -39.887* -33.529* -20.248 (0.631) (20.859) (14.501) (20.035) -0.203 1.094 -21.039* 17.805 (0.271) (12.510) (7.156) (12.40?) 0.098 -6.487* -13.316* 6.590* (0.077) (3.433) (2.092) (3.372) 0.012 -13.652* -10.178* -4.791* (0.028) (1.370) (0.889) (1.316) - a 0.113* 0.084* 0.043* ( a ) (0.014) (0.010) (0.013) -0.222 -0.837 -9.982 -8.985 (0.265) (12.181) (6.742) (11.969) - a 0.001 a * -0.001 ( a ) ( 8 ) ( a ) ( 8 ) a - a - a a ( a ) ( a ) ( a ) ( a ) 0.0121 0.1276 0.1140 0.0800 0.3376 0.2693 0.0524 0.3920 0.0024 0.0001 0.0166 0.0001 0.1263 0.0001 0.3608 0.0001 uls uls uls uls Variable Intercept Height Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq weekend Income Incomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type ALL 533. (125 0. (0. 9. (1 3. <14. -17. <12. 18. <14. 15 (13. 19. (15 -104. <18. -72. <53. 54. <39. -20. <32. 270. <14. -12. <12. -20. (25 -22. <14. -7. <4. -9. (1 0. (0. 9. (14. 0. ODD PHOSPHORUS FAFH (MILLIGRAMS) 919* .236) 563* 181) 651* .929) 088 943) 044 454) 152 626) .678 535) 352 .270) 480* 725) 612 865) 709 042) 090 177) 870* 885) 889 189) 477 .546) 058 568) 521* 020) 465* .561) 068* 016) 156 242) Dill!) 1747 1771 .4719 0001 wls 519. <94. 0. <0. 0. <1 4. <11 -5. <8. -3. <11 -12. <9. -31 <10. 5 <15 8. <38. 71 <38. -42. <26. 8o. <11. 49. <8. -40. <20. -43. (10. .287* .878) .809* .225) .108* .013) .786 .684) .001* 0.1 144* 958) 509* 137) 233 .470) 125 .320) 113 990) 360 .000) 765 632) .422* 966) .362 .068) 787 727) .624* 559) 534 418) 480* 180) 063* 856) 325* 548) 369* 430) a ) a a ) 211 0.6421 0.0287 0.3129 wls FAH 269 (127 0 (0 6 (1 0. (15 -12. (13. 34. (15 27 <14. 49. <15. -80. <19. -F Urbanization Region Race Regression Type ALL QQQ ZINC FAFH (MILLIGRAMS) .904* .064) .011* .003) .053* .031) .133 .246) .304 .206) .372 .245) .321 .224) .482* .250) .471 .311) .234 .921) .776 .596) .209 .507) .155* .240) .017 .203) .218 .407) .964* .249) .023 .065) .033 .025) a a ) .749* .230) a 8 ) a a ) .0831 .1136 .1940 .2337 ols 3. <1 0. <0. 0. <0. -0 <0. -0. 0.0 0.8 0.0 0.2 095* .515) 006* 002) 029 023) .075 178) 096 .148) .211 .174) .175 .157) .568* .178) .180 .234) .949 .682) .606 .479) .423 .424) .802* .172) .455* .149) .146 .344) .633* .187) .238* .047) .133* .019) .001* a ) .467* .155) a a ) a a ) 488 065 169 805 ols FAH 0.0562 0.1443 0.7974 0.3459 ols ALL COPP FAFH ER (MILLIGRAMS) .121 .172) 3* a ) .011* .003) .020 .020) .001 .016) .019 .019) .021 .017) .065* .020) .074* .026) .056 .076) .099 .068) .016 .053) .250* .021) .008 .015) .077* .036) .043* .019) .019* .005) .007* .002) at a ) .011 .018) a!‘ a ) a-k a ) 0.0900 0.4070 0.0009 0.0129 wls 53 0 (0 ( 0. <0. 0. <0. 0. <0. -0. <0. -0. <0. -0. <0. 0. <0. 0. <0. 0. <0. -0 <0. 0. <0. 0. <0. -0. <0. -0. <0. -0. <0. -0. <0. 0. 0. 0. 0. .185 .121) a * 8 ) 003* 002) 024* 014) 021* 011) 012 013) 011 012) 022 013) 025 020) 059 054) 150* 037) .069* 032) 066* 013) 045* 011) 035 026) 071* 014) 023* 003) 009* 002) a * 8 ) .036* .013) a i‘ ) UNI!) ) 0616 1037 4464 0006 uls FAH .031 .167) a ) .005* .003) .025 .019) .012 .015) .056* .018) .010 .017) .022 .019) .025 .024) .035 .074) .091 .069) .075 .053) .235* .020) .103* .015) .094* .036) .022 .020) .008 .005) .014* .002) 8 ) .033* .018) 0.0943 0.4166 0.0043 0.3827 uls Variable Intercept Height Height Urban1 Urban2 Region1 Region2 Region4 Race2 Race3 Race4 Hisp1 Sex1 Employ1 Fstamp1 diet1 Hsize Age Agesq weekend Income Incomesq Imratio* Adj R-SQ F-Tests Prob >F Urbanization Region Race Regression Type ALL 08E 2.1.05 I SODIUM FAFH FAH (MILLIGRAMS) 1174.688*1520. (331.398)(295. 1.983* 1 (0.499) (0. 21.376* -1 (5.105) (4. -103.935* -39. (39.663) (34. —18.992 -22. (33.858) (28. -7.394 -57. (40.427) (33. 16.039 -44. (37.229) (30. -158.734* -95 (39.995) (34. -77.105 -6. (51.179) (44. 432.374* 40. (146.051)(127. 98.096 93. (105.151)(107. -136.649 -110. (87.596) (89. 1084.163* 243 (40.634) (34. 38.379 119. (33.163) (27. 48.891 -57. (72.024) (63. -254.186*-119. (38.244) (32. 36.353* -67. (11.08?) (8. -15.716* -31 (4.186) (3. 0.114* 0. (0.043) (0. 49.749 12. (37.766) (29. -0.002* 0. (0.001) (0. a _ ( a ) ( 0.2714 0.0224 0.0001 0.0065 wls uls .729* -65 0.4983 0.0388 0.8286 122* 122. 862)(363. .556* 0. 429) (0. .048 20. 606) (5. 911 -90. 441) (40. 221 -3. 361) (35. 944* 59. 637) (41 289 61 884) (38. 038) (41 886 20. 291) (53. 122 531 879)(154. 966 136. 172)(110. 022 8. 083) (88. .653* 917. 460) (43. 774*-170. 799) (33 034 107. 337) (72. 619*-150. 717) (38. 443* 111 410) (11 .221* 8. 751) (4. 272* -0. 040) (0. 558 -28. 419) (38. 003* -0. 001) (0. a * 6 ) ( 0.0879 0.1 ALL 093 531. 006><267. 0. <0. 20. <4. -16. .444) .063 <26. 894* 518) 772* 636) 251* 928) 014 176) 476 .598) .869 404) .758 .031) 752*- 473) .192, <31 -25 105 (31 (31 921)(111 930 510) 454 826) 059* 383) 769* .897) 696 631) 434* 256) .969* .395) 622* 246) 061 043) 713 828) 008* 001) a i’ 8 ) 917 0.0427 0.0101 0.0048 uls 225 (81 -61 (31 (25 -93. (52. 20. (31 -27. <0. .660* .230) .000* .033) .043 .059) .002* .001) POTASSIUM FAFH FAH (MILLIGRAMS) 328* 867.526* 290><207.226><276. -0. <0. 441 384) 554* 134) 553 374) .605* .260) 43. <20. 19. .753) 323. <30. -09. .270) .960* .439) .772 <66. 543. 838* 735) 913 589 688) 809 943) 110* .403) -70. .764) 065* 252* 740) 273 .181) 264* 270) 8 a ) 0.1568 0.6356 0.0074 0.0001 uls 0. <0. 1 <3. -0. <24. -20 <19. 2. <23. -10 <21 -40 <24. -9. <32. -6. <73. 103. <00. -70. <50. 177. <24. 90. <19. -22. <43. -77. <23. 817* 295) .387 215) 656 356) .961 876) 111 645) .705 .171) .965* 329) 496 - 158) 250 328)(118. 172. (84. 580* 213) 849 610) 66 15 <4. -20. <32. -6. <27. 127. <32. 55 (30. 63 (33. 265 (39. -72 41 (70. .719 852) 064 401) .934* 289) 851 947) 924 890) 576* 599) .438* 229) .012* 401) .038* 313) .275 400) 659* 937) .234 525) 648* 437.916* 019) 223*- 587) 982 567) 283* 131) .994* .154) .360* .646) .177* .028) .865 .600) .002* .001) a 8 ) 0.0859 0.2441 0.3039 0.1 420 wls (32. 224. (26 -52. <53. 04. <33. 25 <0. 27. <3. -0. <0. -99. <30. -0 <0 ( 920) 984* .887) 381 044) 073* 114) .516* 530) 186* 365) 216* 035) 487* 387) .002* .001) a a ) 0.1169 0.8152 0.0015 0.0001 wls [Blank Page in Original Bulletin] " >- u.» Mention of a trademark or a proprietary product does not constitute a guarantee or a warranty of the product by the Texas Agricultural Experiment Station and does not imply its approval to the exclusion of other products that also may be suitable. All programs and information of the Texas Agricultural Experiment Station are available to everyone without regard " to race, color, religion, sex, age, handicap, or national origin. " M? Copies printed: 1,500