Objectives. To determine the proportion of youth meeting national recommendations for food group intake and to identify food intake patterns.
Design. The US Department of Agriculture's 1989–1991 Continuing Surveys of Food Intakes by Individuals were used to estimate food intake. Intake was determined from 3 days of diet by disaggregating foods into their component ingredients and using weights that correspond to servings.
Participants. The sample included 3307 youth, 2 to 19 years of age, living in the 48 conterminous United States.
Main Outcome Measures. Mean number of servings and percentage of individuals meeting national recommendations for food group intake according to demographic characteristics, patterns of intake, and nutrient profiles associated with each pattern.
Results. Mean numbers of servings per day were below minimum recommendations for all food groups except the dairy group (ages 2 to 11). Percentages of youth meeting recommendations ranged from ∼30% for fruit, grain, meat, and dairy to 36% for vegetables. Sixteen percent of youth did not meet any recommendations, and 1% met all recommendations. The pattern of meeting all recommendations resulted in nutrient intakes above the recommended dietary allowances and was high in fat. Conversely, meeting none of the recommendations resulted in intakes well below the recommended dietary allowances for some nutrients. Total fat and added sugars averaged 35% and 15% of energy, respectively, and levels were similar among most demographic groups.
Conclusion. Children and teens in the United States follow eating patterns that do not meet national recommendations. Nutrition education and intervention are needed among US children.
Recent evidence suggests that dietary intake patterns during childhood and adolescence may predict the occurrence of obesity and adult cardiovascular disease1,2 and may also determine risk for adult diet-related cancers.3 Because of the importance of nutrition in the prevention of disease as well as in growth and development, national recommendations for food intake have been set and are outlined in the Food Guide Pyramid.4 The Pyramid provides recommended intakes for five food groups as shown in Table 1. In addition, national nutrition objectives based on Pyramid recommendations have been determined as part of the Healthy People 2000 objectives.5 Objectives for food intake are to “increase complex carbohydrate and fiber-containing foods in the diets of adults to five or more daily servings for vegetables (including legumes) and fruits, and to six or more daily servings for grain products.” Although this objective was directed toward adults, it was concluded that children and young adolescents “should progress to this type of pattern as well.” Nutrient-specific objectives aim to “increase calcium intake so that at least 50% of youth aged 12–24 … consume daily foods rich in calcium” and to “reduce dietary fat to an average of 30% of calories or less … for people age 2 and older.”
Although data on macro- and micronutrient intake are available for monitoring how well children are eating with respect to nutrient recommendations,6-9 information comparing food intake to food recommendations is limited. The few available reports of food intake are limited to grams of foods consumed, not servings per day of food groups. Because food group recommendations are made in terms of servings per day, this has made comparisons of intakes relative to recommendations difficult. One study examined the number of servings of fruits and vegetables consumed by US children and made comparisons with recent recommendations.10 There have been no reports on children's intake in terms of servings of the grain, dairy, or meat groups, or on patterns of food intake relative to the Pyramid recommendations. Therefore, the purpose of this study was to determine the extent to which children are meeting the Pyramid recommendations, to identify major food intake patterns, and to assess nutrient intakes associated with those patterns.
SUBJECTS AND METHODS
Study Sample and Data Collection
Food intakes of children 2 to 19 years of age were examined using the US Department of Agriculture's 1989–1991 Continuing Surveys of Food Intakes by Individuals. In 1989, 1990, and 1991, the Continuing Surveys of Food Intakes by Individuals sampled housing units in the 48 conterminous United States and the District of Columbia.11 In addition to a basic, all-income sample, an independent sample of low-income households was drawn to support estimates for that group. The low-income sample included only households with a gross income <131% of the federally established poverty threshold. The two samples were combined to maximize sample size. The household response rate for the combined sample was 67% and, coincidentally, 67% of the eligible individuals in these households provided 3 days of food intake data. The sample size for this study was 3307 children 2 to 19 years of age. Table2 shows the characteristics of the respondents in the sample according to subgroups used in this analysis. Weighting factors developed to make estimates more representative of the US population by adjusting for nonresponse and for differential rates of selection were applied in this study.
Although health status indicators were not measured in this survey, the sample can be expected to have similar proportions of healthy children and those with various childhood illnesses, as one would find in any given 3-day period among the general population.
Collection of Food Intake Data
Trained interviewers collected dietary information using a 24-hour recall and then taught the respondent how to record food intake for the following 2 days. The interviewer returned to review and collect the records. Food intake for household members <12 years of age was reported by a knowledgeable adult (in most cases, the primary meal planner). Additional information regarding sampling and dietary data collection procedures is detailed elsewhere.11
Assessment of Food Group Intake
All foods were categorized into groups based on a method developed by the US Department of Agriculture for disaggregating mixtures into their ingredients and assigning those ingredients into Pyramid groups.12 For example, if a meat and vegetable casserole was eaten, meat contributed to the number of servings of meat, vegetables contributed to the servings of vegetables, and so forth. Foods that were not consumed as mixtures were assigned directly to their appropriate Pyramid group.
Gram weights for each food were converted into number of food group servings based on the methods of Cleveland et al.12 For individuals >3 years of age and for 2- to 3-year-olds who consumed >6.7 mJ (1600 kcal), a serving size was defined according to the Pyramid as follows: fruit—half cup of cooked or chopped fruit, quarter cup dried fruit, three-quarters cup of juice, one medium piece of fruit; vegetables—half cup of cooked vegetables, 1 cup of raw leafy vegetables, three-quarters cup of vegetable juice; grain—one slice of bread, one ounce of ready-to-eat cereal, half cup of cooked rice or pasta or amounts of other grain products such as pretzels and cakes that contain an amount of grain equivalent to that in a standard slice of bread; dairy—1 cup of milk or yogurt, 1.5 oz of natural cheese, 2 oz of processed cheeses.
For children 2 to 3 years of age who consumed <6.7 mJ (1600 kcal), serving sizes were set at two-thirds the size of the above-mentioned values for all groups except dairy.4 Weights of foods or ingredients consumed were divided by their respective gram weights per serving to determine the number of servings consumed. Meat and meat alternates were an exception, because they were measured in terms of ounces of cooked lean meat or the amount of meat alternative that was equivalent to the amount of lean meat.
Fat and Sugar
When the Pyramid was developed, nutrient profiles were established for each food group based on typical selections, consumed in their lowest fat and sugar forms (eg, plain breads and cereals rather than pastries, plain vegetables rather than buttered, skim milk, and so on).2,4 Although it was recognized that higher fat and sugar selections might be made from each group, it was felt that the additional fat and sugar in such foods must be accounted for, just as fat and sugar added at the table would be. Thus, the fat and sugar allowance provided for in the Pyramid and depicted by its tip includes all discretionary fat, both naturally occurring and that added in food preparation, and all added sugar.
Consistent with that philosophy, discretionary fat was calculated in this study as the difference between the amount of fat actually consumed and the amount that would have been consumed had the same number of servings of each food group been selected but the leanest choices made and no fat added in cooking or at the table.12It includes the fat from whole milk and nonlean cuts of meat as well as the fat added to food in cooking and fat added at the table in the form of spreads, sauces, and salad dressings. Added sugars were defined as sugars from caloric sweeteners such as those added to breads, cakes, sodas, jams, and ice cream as well as sugars consumed separately. Sugars that occur naturally in foods, such as from milk and fruit, were not counted as added sugars.
Comparison to Pyramid Recommendations and Assessment of Food Intake Patterns
The Pyramid suggests a range of servings for each food group to cover the varying nutrient needs in the population. Unlike recommendations for the other food groups, which are based on energy intake, recommendations for the dairy group differ according to age (two servings for children 2 to 11 years of age and three servings for those 12 to 19 years of age). Table 1 shows the suggested number of servings from each of the other food groups for different energy levels.4
In this study, individual intakes were compared with the upper end of the recommended range for grains, fruits, vegetables, and meats whenever energy intakes were >11.7 mJ (2800 kcal), the middle of the range whenever energy intakes were between 9.7 mJ (2200 kcal) and 11.7 mJ, and the lower end of the range if they were <9.7 mJ (2200 kcal). One exception to this was intakes of 2- to 3-year-olds with energy intakes <6.7 mJ (1600 kcal). For grains, fruits, and vegetables, such intakes were compared with the minimum recommendations, but, as mentioned above, smaller quantities counted as a serving; for meats, the minimum recommendation was lowered to 3 oz (∼66% of the minimum recommendation).
The mean number of servings of each food group was calculated to determine whether children were meeting minimum Pyramid recommendations. In addition, each child's intake of each food group was evaluated according to whether he/she met his/her age- or energy-specific recommendation. The percentage of children meeting the recommendation for each food group was determined, and patterns of intake were ascribed to each individual based on his/her combination of food group recommendations that were met. Thirty-two patterns were possible, ranging from meeting all five recommendations to meeting none. The distribution of all children among the 32 patterns and that of each age, sex, poverty, and race/ethnic group was determined. This method of food patterning is fully described elsewhere.13
Nutrient intakes associated with each pattern of food consumption were determined. Nutrients examined included those with a recommended dietary allowance14 and were also targeted by the Third Report on Nutrition Monitoring Report in the United States.15 Micronutrient intakes were reported as a percentage of the age- and sex-specific recommended daily allowance (RDA). Dietary fiber intake was also examined, although no federal recommendations for fiber are available for children. Total fat, discretionary fat, and added sugars were reported as percentages of energy intake.
Statistical analysis was performed using SUDAAN software (Research Triangle Institute, Research Triangle Park, NC) to account for the complex sample design and weighting factors.16Comparisons of total fat, discretionary fat, added sugars, mean intakes of food groups, and proportion of children meeting Pyramid recommendations between each age, sex, and race/ethnic group were performed using the RLOGIST procedure in SUDAAN. Results are reported for significance levels of P ≤ .01 andP ≤ .05.
Mean Daily Servings From Major Food Groups
Mean height for boys and girls was 54.4 ± 0.6 and 53.9 ± 0.4 in, respectively, and mean body weight for boys and girls was 89.6 ± 2.1 and 86.4 ± 1.9 kg, respectively.
Mean food group intakes for the youth population as a whole were below even minimum recommendations for all food groups except the dairy group (Table 2). Average intakes of milk-based foods by 2- to 5-year-old and 6- to 11-year-old males and females were at or very near recommendations; their intakes of all other food groups were below recommendations. Teen-age males' average intakes met minimum recommendations for grains, vegetables, and meat, whereas teen-age females average intakes did not meet minimum recommendations for any group. The number of servings from the vegetable and meat groups rose, whereas those from the fruit group declined, with age for both females and males.
Of the three racial/ethnic groups examined, whites had the highest number of servings of grains and dairy, whereas blacks had the highest number of servings of meat and the lowest intakes of dairy foods. Intakes of fruit and dairy foods increased with increasing income.
Percentage of Persons Meeting Recommendations
The percentage of children meeting the recommendations for each food group ranged from ∼30% for fruit, grain, dairy and meat, to ∼36% for vegetables (Table 3). For both males and females, the percentage of children meeting the vegetable recommendation rose with age, whereas the percentage meeting the fruit recommendation declined. Teen-age males were more likely than teen-age females to meet the grain, dairy, and meat recommendations.
White children were more likely than black or Hispanic children to meet the recommendation for grains and dairy, but less likely than black children to meet the recommendation for vegetables. The percentage of children meeting the fruit and dairy recommendation was higher in the higher income categories.
Mean Intakes of Energy, Fat, and Added Sugars
Average fat intakes, as a percentage of energy intakes, averaged 35% and did not vary among age, sex, racial/ethnic, or income groups (Table 4). Discretionary fat and added sugars were extremely high, together supplying 40% of the energy in children's diets. This was true for all groups studied. Food energy intake increased with increasing income.
Eighty-seven percent (girls) and 71% (boys) consumed <2200 kcal, 10.5% (girls) and 20.0% (boys) consumed 2200 to 2799 kcal, and only 2.6% (girls) and 9.0%(boys) consumed >2800 kcal.
Distribution of Individuals Across Eating Patterns, by Age and Sex
Table 5 shows the distribution of children in the total population, and in each sex/age group, across the 32 possible dietary patterns. The most outstanding pattern was not meeting any food group recommendations, accounting for 16% of the population; it was the most prevalent pattern for every sex/age group and for every group defined by race/ethnicity or income (data not shown). Only four other patterns—meeting solely the grain, vegetable, fruit, and meat recommendations—represented >5% of the population. These patterns were also dominant for nearly all groups examined (percentages not shown for racial/ethnic and income groups). Together, the patterns of meeting none or only one recommendation accounted for >45% of the children and adolescents in this country. Only 1% of children met all the recommendations, and only 5% met the recommendation for four or more food groups.
Nutrient Intakes Associated With Selected Patterns
Table 6 shows selected nutrient levels for the five most prevalent patterns and the pattern of meeting all the recommendations. The pattern of meeting none of the recommendations had average micronutrient intakes well below the RDA for vitamin B6, calcium, iron, and zinc and the lowest fiber intake of any pattern. Conversely, the pattern of meeting all the recommendations resulted in micronutrient intakes above their respective RDAs for all micronutrients studied, and fiber intakes well above the 11.7 g average for youths. However, fat intakes as a percentage of energy were also high. Youths who met only the grain, vegetable, or fruit recommendation had especially low intakes of calcium and zinc.
Fiber intake was generally low among this population of children. The lowest fiber intake was found in the pattern of not meeting any recommendations (7.6 g), whereas the highest intakes were associated with meeting all the recommendations (21.6 g) or meeting the grain, vegetable, dairy, and meat (21.1 g) recommendations. Of the 32 patterns, 7 resulted in fat intakes that were <32% of energy (none were <30% of energy). In common to these patterns was meeting the recommendation for fruit.
In this study, we found that the intakes of total fat, discretionary fat, and added sugars by US children are well above recommended levels. In fact, discretionary fat and added sugar make up 40% of total energy intake. Additionally, only ∼1% of children are meeting national recommendations for Pyramid groups. We also found differences in meeting Pyramid recommendations between males and females as well as among some groups defined by race/ethnicity and poverty level.
We found that the most prevalent pattern of food intake was not meeting any recommendations and that the next four most prevalent patterns met the recommendation for only one food group. The distribution of intake patterns in children in our study differs from that observed among adults.13 For example, only 9% of adults did not meet any recommendations, and three of the six most prevalent patterns met the recommendations for more than one food group.
All food intake patterns, including the pattern of meeting all the Pyramid recommendations, had high percentages of energy from total and discretionary fats and exceeded amounts of added sugars in sample meal patterns based on the Pyramid.4 These findings also differ from those found in adults; that is, when all the food group recommendations were met, the percentages of energy from fat and added sugar were near recommended levels. In fact, compared with all other patterns, the pattern of meeting all the recommendations resulted in the highest percentage of energy from fat. This finding may be partially attributable to discretionary fat from milk, because nearly 50% of milk consumption by children in this study was from whole milk (data not shown). Therefore, even slight dietary modification from whole to lower fat skim milk would decrease dietary fat intakes among some children. Intake of added sugars was also above recommended levels for all but one pattern, and the pattern of meeting none of the recommendations had the highest percent of energy as added sugars. Our findings emphasize the differences in food choices made within food groups by children as compared with their parents, and points to the need for modification of food choices by children.
Mean servings per day from all Pyramid groups and the percent of children meeting Pyramid recommendations were low for most age, sex, race, and poverty groups. Of interest is the consistent finding that males 12 to 19 years of age were more likely to meet energy-based recommendations for all food groups; had higher average intakes of grains, vegetables, dairy, and meats compared with other age and gender groups; and were least likely to fall into the pattern of not meeting any recommendations. The one area where it appears that females and younger children are doing better than males 12 to 19 years of age is in fruit consumption. Although this finding is positive because these groups are consuming greater amounts of fruits, it also may point to the substitution of lower calorie foods among young females for the purpose of dieting. Indeed, in this study we found very low energy intakes in young females (12 to 19 years of age), which may reflect dieting practices.
Of concern are the food intake differences we found between children based on race. Two of the targeted objectives for the Healthy People 2000 are to increase the intake of grains and calcium-rich foods. It appears that white children may be coming closer to meeting these objectives than are children from other race/ethnic groups. Our findings with respect to food group intake are similar to those in a similar study among adults in the United States.12 For example, in both children and adults, grain and dairy consumption was higher among whites compared with other race/ethnic groups.
The impact of poverty on food intake in this study cannot be ignored. As in other studies,8 lower energy intakes were found among children from less affluent households. In addition, children from less affluent households more often did not meet the recommendations for fruit and dairy intake. The poverty-based findings for children are similar to those found among adults for fruit intake.13
Our finding of low-fiber intakes is similar to that of others13,17 and indicates the need for increased efforts to encourage young children to consume higher fiber foods. As demonstrated by our findings, it is possible for children to consume greater amounts of fiber by following patterns that meet the recommendation for the grains, vegetables, and fruit groups.
The methodologic issue of underreporting that is believed to occur in dietary studies using 24-hour recalls or food records must be addressed when interpreting the findings of this study. To evaluate the potential for underreporting, we used a common method to identify “underreporters” that is based on the Schofield formula for basal metabolic rate19 and using a cutoff of 85% basal metabolic rate developed by Goldberg20 (data not shown). Using this method, we found that ∼10% of children <12 years of age and 35% to 45% of males and females 12 to 19 years of age could be classified as underreporters. However, we did not find differences in the mean intake of food groups, percent meeting recommendations, or differences in food intake patterns either after removing underreporters from the analysis or after adjusting for reported low energy intake.
Another methodologic issue for children is accurate reporting of dietary intake. One study documents overestimates of energy in young males,21 whereas other studies indicate that 24-hour recalls approximate energy intakes of children.22,23
The results of this study indicate that the diets of children in the United States greatly exceed recommendations for fat and added sugars, while at the same time, children are not meeting Pyramid recommendations especially for fruit, grain, and dairy. The data support the need to emphasize low fat, low added sugar options when helping children make food choices, such as encouraging consumption of fruit, vegetables, and grains. There is an immediate need for nutrition intervention and education, especially in adolescent females, minorities, and children from low-income households.
Future research using these data might explore the effect of current recommendations to achieve major dietary goals such as reducing percentage of energy from fat by substituting whole milk with skim milk or replacing high sugar and high fat snack foods with fruits, vegetables, and grains that do not have added fat or sugar. Such research might help to identify which dietary modifications are most likely to be effective in achieving recommendations for foods and nutrients.
- Received September 16, 1996.
- Accepted February 14, 1997.
Reprint requests to (K.A.M.) Merck and Company, Inc, 10 Sentry Pkwy, BL2–3, Blue Bell, PA 19422.
- USDA =
- US Department of Agriculture •
- RDA =
- recommended dietary allowance
- Kemm JR
- Nicklas TA,
- Farris RP,
- Smoak CG,
- et al.
- ↵US Department of Agriculture. The Food Guide Pyramid. Washington, DC: US GPO; 1992
- ↵US Department of Agriculture. Healthy People 2000 National Health Promotion and Disease Prevention Objectives. Washington, DC: US GPO; 1995
- Nicklas TA,
- Arbeit ML,
- Srinivasan SR,
- Berenson GS
- ↵US Department of Agriculture. Food and Nutrient Intakes by Individuals in the United States, 1 Day, 1989–1991. NFS Report No. 91-2. Washington, CD: US GPO; 1995
- Cleveland LE,
- Cook DA,
- Krebs-Smith SM,
- Friday J
- Krebs-Smith SM,
- Ballard-Barbash R,
- Cleveland LE,
- Cook DA,
- Kahle LL
- ↵National Research Council. Recommended Dietary Allowances. 10th ed. Washington, DC: National Academy Press; 1989
- ↵Federation of American Societies of Experimental Biology. Life Science Research Office. Prepared for the Interagency Board for Nutrition Monitoring and Related Research. Third Report on Nutrition Monitoring in the United States: 1995
- ↵Research Triangle Institute. SUDAAN: Survey Data Analysis Software, Release 6.34. Research Triangle Park, NC: Research Triangle Institute, 1993
- Copyright © 1997 American Academy of Pediatrics