The 2005 Dietary Guidelines for Americans indicated that the nutrients of concern for children were vitamin E, calcium, magnesium, potassium, and dietary fiber . Beverages are a potential source of calcium, magnesium, and potassium, hence these nutrients were evaluated in our study, along with phosphorus, a non-fortified nutrient provided by milk. After our analysis was completed, the 2010 Dietary Guidelines for Americans were released, which noted vitamin D, calcium, potassium, and dietary fiber were nutrients of concern for children above the age of 2 . Milk is the primary source of vitamin D in children’s diets, but it was not possible to evaluate vitamin D trends because vitamin D intake was not assessed in NHANES prior to 2002.
Milk has been and still is the beverage consumed by most preschool children. However, there was a trend towards fewer children consuming milk in the last decade as evidence by this data and previous research . Only 77% of children <1–5 years consumed milk of any type during NHANES 2001–2006 compared to 84-85% in NHANES 1988–1994 and NHANES 1976–1980. To ensure optimal growth and health, the 2010 Dietary Guidelines for Americans recommends children aged 2–3 years drink 2 cups (16 oz) and children aged 4–8 years drink 2 Â½ cups (20 fl oz) of fat-free or low-fat milk or equivalent dairy servings a day . On average 2–3 year olds in NHANES 2001–2006 consumed the recommended amount of milk daily. However, children 4 and 5 years of age were consuming 14–15 fl oz/day; to meet the Dietary Guidelines, recommendation through fluid milk intake alone, 20 fl oz/day would need to be consumed for this age group. Data from Krebs-Smith, et al. indicate about one out of four young children 2–8 years old do not meet milk group recommendations . Although the Dietary Guidelines recommend everyone 2 years of age or older consume low-fat (i.e., 1%) or fat-free milk; the majority of milk currently consumed by 2–5 year olds was whole or 2% milk. Several studies have shown potential health benefits of starting consumption of low-fat/fat-free milk early in life [24, 25] and reduced-fat milk would be appropriate for overweight or obese children between 12 months and 2 years of age or those with a family history of obesity, dyslipidemia, or cardiovascular disease . Milk consumption has been reported to increase the likelihood of achieving the recommended intake of calcium by 37% and increase intakes of vitamin A, folate, vitamin B12, and magnesium in 2–5 year olds participating in CSFII 1994–96 . This study found that in addition to calcium and magnesium, milk is a major contributor to the daily intake of phosphorus and potassium in young children. Research has shown young children who drink flavored milk have comparable or higher intakes of many of these nutrients compared to those who exclusively drink plain milk, in part because flavored milk consumers have higher total milk intakes .
Over 50% of children 1–5 years consumed fruit juice in NHANES 2001–2006, representing an increase of 23% from the previous decade. Wang et al.  found a non-significant increase from 41% to 46% from NHANES 1988–1994 to NHANES 1999–2004 in their study of children 2–5 years of age. Fruit juice was an important source of potassium and magnesium in the diets of these children and has been positively linked to achieving recommended intakes of vitamin C and folate in 2–5 year old children in previous research . Mean daily fruit juice consumption in 2001–2006 was 5.4 fl oz/day for children less than 1 year and 10–12 fl oz/day for children 1–5 years. For the entire group, mean intake was 10.3 fl oz/day in NHANES 2001–2006 compared to 9.5 and 7.4 fl oz/day in NHANES 1988–1994 and NHANES 1976–1980, respectively. Consistent with our findings, Wang et al.  observed a significant increase from 9.9 fl oz/day in NHANES 1988–1994 to 11.1 fl oz/day in NHANES 1999–2004 in their preschool sample. This is higher than the fruit juice amount recommended by the AAP, American Heart Association, and Dietary Guidelines (4–6 fl oz/day), and some children were introduced to fruit juice or drinks before the recommended 6 months of age [5, 6, 28].
In the last three decades, at least 35% of young children consumed fruit drinks, which is similar to the percentage reported by other studies [14, 20]. The data indicate a significant increase in the amount of fruit drinks consumed in the last decade, particularly in children 1 and 2 years of age. Children less than 1 year consumed 8.8 fl oz/day and children 1–5 years consumed 10–13 fl oz/day in this analysis of NHANES 2001–2006, which was higher than estimates in CSFII 1994–1996, 1998 (4 fl oz/day) for children 6 months to 6 years  and 8 fl oz/day in CSFII 1994–1996 for children 2–5 years. The contribution of calcium, phosphorus, magnesium, and potassium from fruit drinks was negligible. Generally, the amount of fruit drinks consumed was greater than fruit juice for children of all ages analyzed, yet the caloric contribution of fruit drink and fruit juice to these children’s diets was similar. Across all three decades, fruit drinks provided considerably more calories in the diets of preschool children than soft drinks. Although fruit drinks were not a source of the nutrients we measured, many fruit drinks are fortified with vitamin C which was positively associated with achieving recommended intakes of vitamin C in the study by Ballew et al. . Most fruit drinks contain 10% or less fruit juice and a substantial amount of added sugars.
Approximately one-third of young children consumed soft drinks in all three NHANES analyses. These results are consistent with some previous from CSFII 1994–1996 , but lower than others including studies of NHANES 1999–2002 (39% 2–5 year olds)  and CSFII 1994–1995 (51%) . As children age from 1 year to 5 years, a greater percentage of the population consumed soft drinks. In NHANES 2001–2006, 19% of 1 year olds consumed soft drinks, which increased to 45% by age 5 years. The mean amount of soft drinks marginally increased from 7 fl oz/day in NHANES 1976–1980 to 7.8 fl oz/day in NHANES 1988–1994 and 7.9 fl oz/day in NHANES 2001–2006. These findings are similar to data from CSFII 1994–1996  but appreciably higher than 2.9 fl oz/day indicated by Rampersaud et al.  for CSFII 1994–1996, 1998. Carbonated beverages were negatively associated with meeting the recommended levels of vitamin A, vitamin C, and calcium .
The cumulative consumption of beverages with added sugar has been inversely linked to overall diet quality and meeting the adequacy of several nutrients . In our analysis of the most recent NHANES, an additional 85 empty calories or 5% of total daily intake was consumed as soft drinks. The decline in milk consumption and the steady intake of nutrient poor beverages such as fruit drinks and soft drinks in our analysis is corroborated by others. Carbonated beverages and added sugar juice drinks were inversely associated with milk intakes in children 1–5 years of age in the Iowa Fluoride Study . Harnack et al.  reported in their analysis of CSFII 1994 that high soft drink consumption appears to displace milk and fruit juice in the diets of preschool aged children. For example, those that consumed 9 fl oz or more soft drinks per day were 3.8 times more likely to consume less than 8 oz of milk per day compared to those that did not drink any soft drinks. Although the issue of displacement of milk by other beverages cannot be specifically addressed in cross-sectional data, evidence from longitudinal studies validates this phenomenon. Fiorito and colleagues  followed non-Hispanic white girls from age 5–15 years starting in 1996 and observed that early differences in carbonated beverage intake were predictive of later carbonated beverage and milk intake as well as selected nutrients. Girls who consumed carbonated beverages at age 5 years had higher intakes of carbonated beverages, lower milk intake, higher intake of added sugars, lower protein, fiber, vitamin D, calcium, magnesium, phosphorous, and potassium for the 10 years of follow-up. Another study of non-Hispanic white girls who were followed from age 5–9 years observed that girls who consumed adequate amounts of calcium consumed on average twice as much milk, had smaller decreases in milk intake, and consumed 18% less sweetened beverages .
The link between beverage consumption, energy intake and body weight have been evaluated in this young population, and is a significant concern because obesity tends to track over time . High intake of sweetened beverages is linked to increased energy intake, weight gain and adiposity in children. Regular consumption of sweetened beverages between meals by children 2.5–4.5 years more than doubled the odds of being overweight at 4.5 years . In a longitudinal study of girls, sweetened beverage consumption (sodas, sport drinks, fruit drinks, and sweetened coffee or tea) at age 5, but not milk or 100% juice, predicted adiposity in childhood and adolescence . Higher intake of sweetened beverages at 5 years was correlated with a higher weight, percent of body fat, and waist circumference from 5–15 years of age . Sweetened beverages in both of these longitudinal studies did not include flavored milk but included sugar-sweetened or artificially sweetened fruit drinks and carbonated beverages. Excessive fruit juice consumption in preschool children has been associated with obesity in some  but not all studies [11, 35]. Wiley  found milk intake was positively associated with BMI among children aged 2–4 years in an analysis of NHANES 1999–2004 data while a cross-sectional analysis of NHANES 1999–2002 data of 2–5 year olds revealed that energy intakes of flavored milk drinkers were higher than those that consumed plain milk or non milk drinkers, but BMI indices did not differ among the three groups . Huh et al.  reported in their study that milk intake at age 2 years, whether full or reduced-fat, was not linked to the risk of overweight at age 3 years.
Our analysis was limited by the use of cross-sectional data and associated analytical data across three time periods. But in the case of nationally, representative dietary data, NHANES data is the only source currently available. Given the cross-sectional approach of NHANES causality cannot be determined. Also the data was based on a single 24-hour dietary recall which may not be a fair representation of an individual’s usual intake. The 24-hour dietary recall has been shown to be prone to both under- an over-reporting. Nevertheless, this method has been shown to be valid in estimating the mean intake of a population [38, 39].
In summary, the number of young children consuming milk has significantly declined while those consuming fruit juice has increased dramatically in the last decade compared to the previous two decades. The proportion of children that consume fruit drinks and soft drinks has remained high and relatively stable across the three decade time period. Milk is the main source of calcium in the diets of these children, which has been declining alongside increased reduced-fat milk consumption. Milk is also a major source of magnesium and potassium, short-fall nutrients identified by the 2005 and 2010 Dietary Guidelines for Americans. The amount of milk that is consumed in this age group is less than recommended, particularly among 4 and 5 year olds, where fruit juice consumption is significantly higher than recommended. Fruit drinks are a significant source of calories. A greater number of children consume soft drinks as they age from 1–5 years. Since dietary patterns established as young children are carried throughout childhood and adolescence, and a link between non-milk sweetened beverages and obesity has been increasingly demonstrated, it is prudent that parents, educators and child caretakers replace some of the nutrient poor beverages young children are currently consuming with low-fat and fat-free milk.