This is the first longitudinal study of girls that investigated both the quality and tracking of nutrient and food group intakes across middle childhood. Our data provides evidence of food and nutrient intakes tracking and declining in quality during middle childhood. Tracking was evaluated using two methods: descriptive plots and percent agreement between quartiles. Both methods provide support for dietary intakes tracking during middle childhood. Only 1% of girls in this study met all of the FGP recommendations at age 5 and none of the girls met all of the recommendations at ages 7 and 9. These findings are similar to those reported by others based on nationally representative data [30
Several researchers have investigated whether children's and adolescents' dietary patterns track over time. Studies have ranged from 1 year to more than 10 years and have included children as young as 6 months, preschool-, elementary- and high-school-aged children, and young adults. Macro- and micronutrients have been examined as well as food groups and food choices. Fruit and vegetable intakes were reported to have low to moderate tracking during elementary school [27
] but showed stability from adolescence into adulthood [23
]. Nutrient intakes were found to track during early and middle childhood [24
] and during young adulthood [31
] but not during adolescence [32
]. Food choices, however, were found to track during adolescence [34
] and from childhood to young adulthood [35
]. Some of these results are conflicting and it is difficult to draw conclusions from these studies due to variations in the ages of the participants, study duration, diet assessment methods, and methods used for the tracking analyses.
Results from the current study provide evidence of nutrient and food group intakes tracking during middle childhood. Among nutrients, energy, protein, cholesterol, vitamins D and E, phosphorus, magnesium, iron, and zinc showed the strongest tracking. Plots of these nutrients showed that each quartile remained distinct over time. The rate of exact agreement between quartiles (highest and lowest) for these nutrients was approximately 45%, which exceeds the rate expected by chance (i.e. 25%). Other micronutrients also showed evidence of tracking, although not as strong. Among food groups, fruit and dairy intakes showed the strongest evidence of tracking. On the other hand, tracking of grain and meat intakes was low.
There were several lines of evidence to indicate that diet quality declined between 5 and 9 years of age. A larger proportion of girls at 9 years of age failed to meet the dietary recommendations for most food groups than they did in the younger years. At age 5, 69% of the girls met the dairy recommendation, however by age 9 only 36% met this recommendation. Few girls met the fruit and vegetable recommendations at age 5 (27% for fruit and 8% for vegetable) and even fewer met these recommendations at age 9 (7% for fruit and 3% for vegetable). It might be argued that decline in dietary quality reflects the increased recommendations for the older girls. The FGP [12
] recommendations specify the number of servings for children aged 2–6 years and older children, therefore at ages 7 and 9 the recommended number of servings is increased. This could explain why fewer girls were meeting recommendations at age 9, however energy intakes were increased at age 9 making it possible for girls to come closer to meeting the increased FGP recommendations.
The data on nutrient density indicate that the decline in quality is not fully accounted for by increases in recommendations that occur at ages 7 and 9. Nutrient densities of several nutrients, including vitamins C and D, calcium, phosphorus and magnesium were lower at both 7 and 9 years of age than at 5 years; zinc density was lower at 9 years as compared to 5 years. While energy intake increased, there were no increases in the number of servings in the dairy or fruit group, which could explain the decreased density for nutrients such as calcium and vitamin C. More comprehensive analyses of food subgroups would be necessary to determine whether choices made within the major food groups were more or less nutrient dense.
Our data show that diet quality decreases but also that dietary intakes track across middle childhood. In other words, girls who eat few servings of dairy at age 5 are likely to eat few servings of dairy at age 9. Likewise, girls who eat greater amounts of dairy at age 5 are likely to eat greater amounts at age 9. From an intervention perspective this suggests that eating patterns are developed early in life, before the age of 5. Thus intervention approaches to reduce the risk of chronic diseases such as osteoporosis would need to be targeted at a very young population and their caregivers. Although, our data provide evidence that diet quality will decline during middle childhood, those who start out with 'healthier' diets at a young age will be less likely to be at risk for inadequate intakes during middle childhood than those who start out with less healthy eating patterns.
Several micronutrients, including thiamin, riboflavin, and niacin, were not reported in this study because analysis of these nutrients revealed that mean intakes from our sample were nearly twice the RDA. The pattern of findings for food group intakes in combination with micronutrient data suggest that adequate intake of these vitamins was most likely a result of food fortification. For all three of these vitamins, fortified ready-to-eat cereals are top contributors to intakes of the U.S. population [14
] and children commonly consume ready-to-eat cereals. Food fortification also explains the increase in nutrient density of folate. During the course of this study The Department of Health and Human Services mandated that all enriched cereal grains be fortified with folate, whereas previously only ready-to-eat and cooked cereals were fortified with folate [14
]. These findings that fortified foods contribute substantially to children's micronutrient intakes are similar to those reported by Berner [36
] and Subar [37
] for nationally representative data.
In this study, mean calcium intakes were below the recommendations and only one-third of girls met the FGP recommendations for dairy at age 9. It has been reported that calcium intakes are declining among children as a result of decreased dairy consumption and that it is not possible to obtain adequate calcium without a source of milk in the diet [38
]. Furthermore, drinking milk during childhood has been associated with the likelihood of meeting calcium recommendations [39
] and with reducing the risk of developing osteoporosis later in life [40
]. These findings emphasize the importance of milk and dairy foods, especially during childhood.
A significant increase in the prevalence of overweight among our sample was noted. By age 9, nearly one-third of our sample was at or above the 85th
BMI-for-age percentile. The prevalence of overweight in our sample of girls was similar to national data, which report 20% of females aged 2–5 years and 28% of females aged 6–11 years are above the 85th
]. When our sample was split into quartiles based on BMI percentiles at age 5 and plotted the results showed nearly perfect tracking (data not shown); girls maintained their relative quartile ranking for BMI percentile and the lines for each quartile were parallel.
This study has some limitations. The sample is exclusively white girls and at entry into the study included only 2-parent families. Thus, our results cannot be generalized to other socioeconomic, ethnic and racial groups. Supplement intake was not quantified; however, the prevalence of supplement use declined from 52% of the sample at age 5 to 32% at age 9. Also, supplement users in this sample of girls were found to have higher quality diets than nonusers [42
], thus it is unlikely that including intakes from supplements would reduce the proportion of girls in this sample who were found to be at risk. Children's dietary data were recalled by mothers and daughters and are subject to errors of under-reporting or over-reporting for individuals, however we used the multiple pass technique to facilitate recall and reduce error. Intra-individual variation in intake does vary by nutrient, however, we collected 3 days of dietary intake data, which is generally acceptable for estimating the usual intakes of groups. Finally, continued dietary monitoring of the girls in this study is important as the girls make the difficult transition from childhood to adolescence.