Changes in Children's Diets:
1989-1991 to 1994-1996

EXECUTIVE SUMMARY

Children's diets may influence their lives in a variety of ways, including affecting their growth, health outcomes, and cognitive development. Since the late 1980s, considerable attention has been given to identifying nutritional problems in children's diets and identifying and implementing initiatives that may help children (and adults) improve their diets. For example, public initiatives have focused on encouraging children to limit their fat intake and to eat a balanced diet that includes recommended numbers of servings of the five major food groups from the U.S. Department of Agriculture's Food Guide Pyramid. As a result, there is wide interest in whether children's diets have changed during recent years.

This report is the second of two reports on the nutrition of children using findings from the analysis of the 1989-1991 and 1994-1996 panels of the Continuing Survey of Food Intakes by Individuals (CSFII). The key objectives of the overall study are to describe the diets of school-aged children in the United States as of the mid-1990s, examine relationships between children's participation in the school meal programs and their dietary intake, and examine changes in intake between the periods 1989-1991 and 1994-1996. This second report describes changes between these periods in children's food and nutrient intakes and reports the percentages of children meeting various dietary standards. The first report (Gleason and Suitor 2000) focuses on children's dietary intakes as of 1994-1996 and also compares the diets of participants and nonparticipants in the school meal programs.

The 1989-1991 and 1994-1996 CSFII surveys collected dietary intake and other data from nationally representative samples of noninstitutionalized residents of the United States. The 1989-1991 CSFIIl collected dietary intake data on three consecutive days, with data on the first day collected through a 24-hour recall interview and data on the second and third days collected through food diary records. In contrast, the 1994-1996 CSFII collected two nonconsecutive days of dietary intake information, both by 24-hour recall interviews. The analysis in this report uses data from more than 2,900 children ages 6 to 18 who completed the first day of the 1989-1991 dietary intake interview and from nearly 2,700 children ages 6 to 18 who completed both days of the 1994-1996 dietary intake interviews.

The analysis presented in this report includes several important methodological features. To address the issue of what proportion of children meet various dietary standards, we used statistical methods to obtain unbiased estimates of the distribution of usual intake using two or three days of intake information for each child. For assessing the nutrient adequacy of children, we used the newly developed Dietary Reference Intake (DRI) values when possible. Since accepted reference standards (Estimated Average Requirements, or EARs) have not yet been developed for nutrients other than the B vitamins, phosphorus, and magnesium, we assigned reference standards derived from the 1989 Recommended Dietary Allowances (RDAs) for these nutrients. No reference standard was assigned for calcium; rather, the change in the full distribution of calcium intake was examined.

To describe changes in children's diets over time, we used data from two different surveys using two different samples. The analysis considered potential methodological differences between the two surveys. For example, because of differences in dietary intake interview methodology, much of the analysis is based on the Day 1 intake data, since both the 1989-1991 and 1994-1996 CSFII collected these data using a dietary recall interview. In particular, we limited the analysis to day one only when comparing mean intakes across the two periods. When comparing the distribution of usual intakes across time periods, however, we used three days of 1989-1991 CSFII data and two days of 1994-1996 CSFII data due to the statistical requirements for estimating the distribution of usual intake. Finally, we conducted tests to determine whether the observed changes in children's diets over time were statistically significant.

Major Findings

Food energy intake among school-aged children increased between 1989-1991 and 1994-1996.

Over the period 1989-1991 to 1994-1996, school-aged children's mean 24-hour food energy intake increased significantly from 88 to 94 percent of the 1989 Recommended Energy Allowance (REA). Comparison of the full distributions of food energy intake for the two periods shows that intakes increased among light, moderate, and heavy eaters. In addition, the increase was driven primarily by foods eaten at dinner and for snacks during the day, as children's mean food energy intakes at breakfast and lunch stayed about the same over this period.

The increase in energy intake held for all school-aged males and for females ages 14 to 18 but did not hold for the younger females. For females ages 6 to 8, in fact, mean food energy intake decreased from 94 to 87 percent of the REA (although this difference was not statistically significant). Among racial/ethnic groups, the increase in food energy intake was largest for white children. Black and Hispanic children experienced smaller (and statistically insignificant) increases in food energy intake.

Despite the increase in food energy intake, children's intakes of most vitamins and minerals did not change much over this period.

While school-aged children consumed more food energy in the middle 1990s than they did in the late 1980s to early 1990s, their intake of most vitamins and minerals stayed about the same. Of the 14 vitamins and minerals examined in this report, children's mean 24-hour intakes of only iron and niacin increased significantly between 1989-1991 and 1994-1996. An alternative way of measuring nutrient intake--estimating the proportion of children whose usual intakes meet recommended dietary standards (the EAR, if available)--tells a similar story. There were significant increases in the percentage of school-aged children meeting the relevant standard for only two of the vitamins and minerals--iron and vitamin E. To some extent, this was due to the fact that very high proportions of children met the standard for most water-soluble vitamins and vitamin A in the earlier period, leaving little room for improvement. However, relatively small percentages of children met the standards for nutrients like folate, magnesium, and zinc in 1989-1991, and these percentages did not increase over time either. The distribution of calcium intake among all children did not change significantly between 1989-1991 and 1994-1996, but there was a decrease in calcium intake among the younger and older females and among black children.

Males were more likely than females to experience increases in mean vitamin and mineral intake between 1989-1991 and 1994-1996. For example, mean intakes of vitamin E, vitamin B12, niacin, iron, and zinc increased over this period among males ages 6 to 8, as did mean intakes of vitamin B6, niacin, iron, and zinc among males ages 9 to 13. Among females, only iron intake increased over this period. Among females ages 6 to 8, in fact, mean intakes of vitamin E, riboflavin, folate, calcium, magnesium, and phosphorus decreased significantly between 1989-1991 and 1994-1996.

The pattern of an increase in food energy intake without a corresponding increase in vitamin and mineral intake is particularly clear among females ages 14 to 18. The mean energy intake among these teenage girls increased from 78 to 86 percent of the REA between 1989-1991 and 1994-1996. Over the same period, mean intake of only one micronutrient (iron) increased significantly. Since the increase in food energy intake clearly was not driven by an increase in foods rich in vitamins and minerals, one possibility is that it was driven by an increase in foods or drinks high in added sugars. This hypothesis is supported by two other trends over this period among females ages 14 to 18: (1) their intake of carbohydrates as a percentage of food energy increased, and (2) their consumption of soda and fruit drinks and fruit-flavored drinks increased. The increase in mean iron intake may have been due to the increased availability and consumption of highly fortified breakfast cereals.

Mean vitamin and mineral intakes increased more among white children than among the other racial/ethnic groups. For example, white children's mean intakes of vitamin B6, niacin, thiamin, iron, and zinc increased significantly. Furthermore, the percentage of white children whose usual intakes met the dietary standard increased for vitamin E, iron, and zinc. Among black and Hispanic children, by contrast, mean intakes of no vitamins and minerals changed significantly over this period. Although the percentage of black children who met the dietary standard for vitamin C intake increased significantly between 1989-1991 and 1994-1996, the percentage meeting the folate standard decreased significantly, from an already low level of 47 percent in the earlier period to 38 percent in the later period. As mentioned above, black children's calcium intake also decreased over this period, especially among those with the highest intakes of calcium in the earlier period.

Children's fat intake as a percentage of food energy decreased between 1989-1991 and 1994-1996, although absolute intakes of fat did not decline.

Among all school-aged children, mean 24-hour total fat intake decreased from 34.1 to 32.4 percent of food energy, while saturated fat intake decreased from 13.0 to 11.7 percent of food energy. These decreases occurred at breakfast and lunch, as well as over 24 hours. Each of the age/gender groups we examined experienced similar decreases in fat intake, but there were differences by race/ethnicity. White children's total fat and saturated fat intake decreased substantially over this period. Among black and Hispanic children, however, total fat intake as a percentage of food energy stayed about the same, and saturated fat intake as a percentage of food energy decreased only moderately.

The decrease in fat intake as a percentage of food energy among all children translated into large increases in the percentage of children meeting the dietary guidelines for total fat and saturated fat intake. In 1989-1991, only 14 percent of children met the dietary guideline of limiting total fat to no more than 30 percent of food energy, and only 7 percent met the guideline of limiting saturated fat to less than 10 percent of food energy. By 1994-1996, these percentages had approximately doubled, to 25 percent for total fat and 16 percent for saturated fat. Despite the large increases in the percentage of children meeting the dietary guidelines for total fat and saturated fat, most children still failed to meet these guidelines by 1994-1996. In addition, the percentage of black and Hispanic children who met the total fat and saturated fat dietary guidelines did not increase over this period.

Despite the decline in fat intake as a percentage of food energy, children's fat intake measured in absolute terms did not change between 1989-1991 and 1994-1996. In the earlier years, children's mean intakes of total fat and saturated fat were 77 grams (g) and 29 g, respectively. In the later years, these mean intakes were about the same (78 g of total fat and 28 g of saturated fat). The decrease in fat intake as a percentage of food energy was caused by an increase in food energy intake without a correspondingly large increase in fat intake.

Children's protein intake as a percentage of food energy decreased between 1989-1991 and 1994-1996, and their carbohydrate intake increased over this same period.

Between 1989-1991 and 1994-1996, children's mean 24-hour intake of protein decreased significantly, from about 15 to 14 percent of food energy. This decrease in protein intake held across the board--at breakfast and lunch, for each of the age/gender groups, and for each of the racial/ethnic groups. However, children's absolute intake of protein did not change significantly, and was well above the protein RDA for most children. With the decreases in the proportion of food energy from both fat and protein intake, there was a corresponding increase in the proportion of food energy from carbohydrates. Children's mean 24-hour carbohydrate intake increased from 52.0 to 54.7 percent of food energy over this period. This was associated with a large increase in the percentage of children meeting the dietary recommendation of consuming more than 55 percent of food energy from carbohydrates (from 28 to 53 percent). Children's mean absolute carbohydrate intake also increased significantly.

Children's cholesterol intake decreased, and their fiber intake increased, between 1989-1991 and 1994-1996.

Over 24 hours, children's mean intake of cholesterol fell from 257 milligrams (mg) in 1989-1991 to 241 mg in 1994-1996. This decrease in cholesterol intake occurred primarily at breakfast, where mean intake fell from 71 to 57 mg. In addition, the decrease was strongest among younger children (under age 14) and white children. Mean cholesterol intake among children ages 14 to 18 and among blacks and Hispanics did not significantly decrease over this period. The percentage of all school-aged children who met the dietary recommendation of limiting their intake to no more than 300 mg increased slightly from 74 to 78 percent.

Among all school-aged children, mean fiber intake increased significantly, from 13.4 to 14.2 g between 1989-1991 and 1994-1996. This increase was driven by foods consumed at meals other than breakfast and lunch (at which fiber intake did not change over this period). The increase in fiber intake was associated with the overall increase in food energy intake; fiber intake per 1,000 calories of energy did not increase significantly. The increase in mean fiber intake was strongest among males and females ages 14 to 18 and among white children. As with cholesterol, the percentage of children meeting the recommended standard for fiber intake remained consistent, but at a low value (28 percent).

Over the period under study, children increased their consumption of vegetables and grain products and consumed fewer milk products and meat and meat substitutes.

Children's mean daily intake of grain products increased from 6.5 servings in 1989-1991 to 7.2 servings in 1994-1996. At the same time, their mean daily intake of vegetables increased from 2.3 to 2.6 servings. This increase in vegetable consumption was responsible for the increase in combined servings of fruit and vegetables, from 3.7 to 4.1 servings, bringing children's average consumption closer to, but still well below, the goal of a minimum of five servings of vegetables and fruit per day. The increase in vegetable consumption was largest among males and females ages 14 to 18.

In addition to the increase in mean consumption of grain products and vegetables, there were increases in the percentage of children whose intake of these foods was above recommended levels. For example, the percentage of children meeting an age/gender-specific target for grain product consumption increased significantly, from 17 to 23 percent. The corresponding increase for vegetable consumption was from 12 to 17 percent. As these figures show, however, most children failed to meet the recommended consumption levels even after the increase in grain product and vegetable consumption.

The patterns for the consumption of milk products and meat and meat substitutes are the reverse of those for grain products and vegetables. Children's mean intake of milk products decreased significantly, from 2.4 to 2.0 servings per day between 1989-1991 and 1994-1996, and their mean intake of meat and meat substitutes decreased significantly, from 1.7 to 1.4 servings per day. The percentage of children meeting the age/gender-specific targets for the consumption of these food groups also decreased significantly over this period, from 40 to 30 percent for milk products and from 19 to 7 percent for meat and meat substitutes.

Children's beverage consumption shifted between 1989-1991 and 1994-1996 from high-fat milk to lower-fat milk, soda, and fruit and fruit-flavored drinks.

The decline in school-aged children's consumption of milk products was driven primarily by a decrease in their consumption of whole milk, from 0.7 to 0.4 servings per day. However, their consumption of low-fat milk remained constant and their consumption of nonfat milk increased slightly but significantly over this period. Children consumed greater amounts of regular and diet soda (mean intake increased from 1.0 to 1.4 servings per day) and fruit drinks and fruit-flavored drinks (mean intake increased from 0.5 to 0.8 servings per day). These changes in beverage consumption were especially pronounced among older children (those ages 14 to 18).

Changes in school-aged children's consumption of specific foods between 1989-1991 and 1994-1996 were consistent to some extent with changes in their nutrient intake.

Several changes in children's food consumption were consistent with the substantial decrease in fat intake as a percentage of food energy. In particular, the declines in the consumption of whole milk and meat, along with the increases in the consumption of nonfat milk, soda, fruit drinks, and fruit-flavored drinks, are a potential explanation for the decrease in fat intake. The decrease in milk intake among females ages 6 to 8 and among black children may also help explain why mean calcium intake decreased significantly for these subgroups. Finally, the increase in children's mean consumption of vegetables and grain products may have played a role in the small increase in their fiber intake between 1989-1991 and 1994-1996.

January 2001

Last modified: 12/04/2008