Assessment of Beverage Trends and Replacing Nondairy Caloric Beverages with Milk at Meals across Childhood Improves Intake of Key Nutrients at Risk of Inadequate Consumption: An NHANES Modeling Study

Background Milk is a key source of important nutrients including the nutrients of public health concern. However, most Americans do not meet current (dairy) United States Department of Agriculture (USDA) dietary guideline recommendations, and the intake has been declining. Objective The aim of this study was to investigate milk and beverage intake trends and nutrient intakes from these products in United States children aged 6–18 y and to model the effect of isocaloric substitution of nondairy beverages with milk. Methods Data from National Health and Nutrition Examination Survey (NHANES) 2001–2018 for children age 6–8 (N = 4696), 9–13 (N = 8117) and 14–18 y (N = 8514) were used with milk and other beverage intakes determined from the first 24-h in-person dietary recall. Nutrient intake was determined using the NHANES cycle-specific total nutrient intake files. Nutrient modeling was performed by isocaloric substitution with milk of all nondairy beverages consumed during lunch and dinner meals combined. Sample-weighted analyses were performed using SAS 9.4. Results Between ages 6–8 and 14–18 y, daily intake of milk and flavored milk decreased by 10% and 62%, respectively, while daily intake of caloric beverages excluding milk increased by 96%. Daily intake from caloric beverages and milk combined decreased for fiber, protein, fat, saturated fat, calcium, magnesium, potassium, vitamin A, and vitamin D and increased for energy, carbohydrates, added sugars, and folate between ages 6–8 y and 14–18 y. Isocaloric substitution of all caloric nondairy beverages at meals with milk (using nutrient contribution of USDA milk, not further specified (NFS)) resulted in increases in protein, fat, saturated fat, calcium, magnesium, potassium, sodium, vitamin A, folate, vitamin B12, and vitamin D and decreases in carbohydrate, fiber, and added sugar. Conclusion These findings provide additional evidence to support dietary recommendations for milk, and efforts should be made on behalf of leading health professionals and childhood meal programs to highlight milk as a beverage of choice in children and adolescents.


Introduction
The Dietary Guidelines for Americans are intended to promote the intake of a variety of food groups, including dairy, and provide dietary advice to help meet nutrient needs, promote health, and reduce risk of disease across all stages of life.The current version recommends 2.5 daily servings (cup equivalents) of dairy products for children aged 4-8 y and 3 daily servings for those aged 9-18 y to ensure that they meet nutrient intake recommendations, promote healthy growth and development, as well as reduce risk of chronic diseases [1].
Cow milk is a key source of high-quality protein and essential micronutrients, such as calcium, phosphorus, vitamin A, vitamin D, riboflavin, niacin, pantothenic acid, vitamin B 12 , iodine, potassium, selenium, and zinc [2] and contributes to overall diet quality in both children and adults [1].Calcium, potassium, dietary fiber, and vitamin D are underconsumed in the general United States population and have been identified as "nutrients of public health concern" because low intakes are associated with health concerns [1].Additionally, children and adolescents underconsume phosphorus, magnesium, and choline among both male and females and protein, iron, folate, vitamin B 6 , and vitamin B 12 among females [3].Data suggest that children and adolescents who meet dairy recommendations are less likely to be below recommended levels (Estimated Average Requirements or Adequate Intakes) of several essential nutrients, including calcium, magnesium, phosphorus, riboflavin, vitamin A, vitamin B 12 , vitamin D, selenium, potassium, and choline [4].Moreover, milk was reported to be the leading food source of 3 of the 4 nutrients of public health concern, calcium, vitamin D, and potassium, providing 22%-23%, 46%-47%, and 12%-13%, respectively, of daily intake for children aged 6-18 y [5].
Approximately 90% of Americans do not meet the recommendations for dairy intake, and only ~65% of children, 34% of adolescents, and 20% of adults drink milk on a given day [1].Additionally, intakes of dairy and particularly fluid milk have been slowly but steadily decreasing in the United States as well as globally [6][7][8][9][10].Daily per capita consumption of fluid cow milk has declined by approximately 50% from approximately a cup in 1970 to about half cup in 2019 [6,7].Although sugar-sweetened beverages (SSBs), such as soft drinks and juice drinks appeared to be replacing milk in earlier years, the current increase in the number of caloric beverages commercially available, such as energy drinks, or more recently, the proliferation of plant-based milk alternatives may help to explain this trend; however, the extent of this impact is yet to be determined [6,7].
The current study investigated the current trends in fluid milk and beverage intake as well as nutrient status in United States children aged 6-18 y using the NHANES, a nationally representative, continuous survey of the noninstitutionalized civilian United States population [11].We hypothesize that as milk consumption declines in this age group, consumption of less nutrient-dense beverages will rise, resulting in a drop of certain essential nutrients in the overall diet and a reduction of nutrient adequacy and overall diet quality.We further hypothesize that adding 1 serving of milk per day or isocaloric substitution of all caloric beverages consumed at meals with milk will result in significant improvement in overall nutritional intake and consequently nutrient adequacy within this age group.

Milk, beverages, and nutrient intake
The milk and beverage intakes were estimated from the first 24-h in-person dietary recall.Data were collected by proxy (the person with the most knowledge of their food intake)-assisted interviews of children aged 6-11 y and individual interviews for those aged !12 y, as per NHANES dietary interview protocols.Details of NHANES protocols are available in previously published literature [11,13].WWEIA food categories were used to define milk and other beverages: milk (USDA subgroup 10), flavored milk (USDA subgroup 12), milk substitutes (USDA category 1404), fruit juice (USDA subgroup 70), SSBs (USDA subgroup 72), soft drinks (USDA category 7202), fruit drinks (USDA category 7204), coffee and tea (USDA subgroup 73), caloric beverages, excluding milk (USDA main group 8 and subgroups not in 71, 77, 78), and caloric beverages, including milk (also included milk, flavored milk, and milk substitutes) [14].Nutrient and energy intake was determined using the NHANES cycle-specific total nutrient intake files based on Food Nutrient Database for Dietary Studies, which provides nutrient content for food and beverages reported in WWEIA [14].

Nutrient modeling
The following 2 modeling scenarios were utilized to assess changes in nutrient intake: 1) 1 serving of milk was added to the daily diet of each subject using nutrients from 1 cup of milk (NHANES food code 11100000; milk, nfs) (nfs, not further specified, which is a combination of various fat levels of milk); and 2) an isocaloric substitution of all nonmilk caloric beverages consumed at meals with milk where nutrients from all caloric beverages (SSBs, milk substitutes, or all beverages in WWEIA beverage subgroup) consumed during lunch and dinner meals combined were substituted with nutrients from isocaloric amount of milk using nutrient profile of NHANES food code 11100000 milk, nfs.Nutrient contents of milk (NHANES food code 11100000; milk, nfs) are provided in Supplemental Table 1.

Statistical analyses
All analyses were performed using SAS 9.4 (SAS Institute) using PROC SURVEYMEANS, and survey parameters, including primary sampling unit, strata, and day 1 dietary weights were used to account for NHANES' complex sample design and to provide nationally representative estimates using Taylor series expansion to generate variance estimates.The nutritional contribution of caloric beverages was assessed using the recommended population ratio method.Regression analyses were used to assess changes in intake across the age range (6-18 y) using PROC SURVEYREG; the regression coefficient generated from these analyses represented the change across age in change per year.Sex, ethnicity, and PIR level were used as covariates.Although some analyses were conducted across the entire age group (i.e., regression analyses examining changes across age and nutrition contribution of caloric beverages), other analyses were also conducted for separate age groups (6-8, 9-13, and 14-18 y) to allow comparisons with previous published results.Because typical statistical testing was not appropriate for modeling of adding a serving milk to all subjects (given violation of assumption of independent samples), we set a 10% change in intake as a meaningful difference.For replacement of caloric beverages at meals, paired t tests were used to assess changes in intakes.

Nutrient contribution of the caloric beverages
Nutrient contributions of the caloric beverages including and excluding milk to daily intakes of children aged 6-18 y are presented in Figure 1.As expected, nutrient contributions of caloric beverages including milk were more than that of caloric beverages without milk.Caloric beverages, including milk, contributed 18% of total calories, and more than half (11%) was provided by caloric beverages excluding milk.Caloric beverages excluding milk contributed to 1%-2% of daily intakes of protein, vitamin A, vitamin B 12 , vitamin D and 5% calcium while caloric beverages, including milk contributed to 13% protein, 24% vitamin A, 24% vitamin B 12 , 45% vitamin D, and 31% calcium.Total fat and saturated fat were mainly contributed by caloric beverages including milk (8% and 12%, respectively) while caloric beverages excluding milk contributed <1% total fat and saturated fat.Caloric beverages including or excluding milk contributed comparable amounts (42% compared with 38%, respectively) of added sugars (Figure 1).

Discussion
The results of the present analysis indicate that with increasing age, i.e., moving from early childhood to teenage years, the intake of milk decreased whereas the intakes of soft drinks and SSBs increased and intakes of milk substitutes, 100% juice, and fruit drinks did not change.Consequently, the intakes of most nutrients from all caloric beverages, including milk decreased with age whereas the intake of energy and sugars (including added sugar) increased with age.The addition of a serving of milk to the daily diet or isocaloric replacement of all caloric beverages at lunch and dinner with milk resulted in an increase in several micronutrients including calcium, vitamin A, vitamin B 12 , and vitamin D with only a 1-3 g/d increase in saturated fat across both modeling scenarios and a 2-3 teaspoon equivalent/d (8-12 g/d) decrease in added sugars for replacing caloric beverages at meals.
Milk is an important of nutrients, and the consumption of milk by children has been positively correlated with diet quality as well as meeting nutrient recommendations, including vitamin A, folate, vitamin B 12 , calcium, and magnesium in the United States [5,16] as well as globally [17][18][19][20][21].In our analysis,  beverages, including milk, provided substantial amounts (>20% of daily intake) of vitamins and minerals while the contribution of beverages excluding milk was minimal ( 5%) to daily intakes of these vitamins and minerals.This suggests that nutrient density (nutrient to energy ratio) of milk is much more than that of other caloric beverages.In the present analysis, the intake of milk decreased and the intakes of soft drinks and SSB increased with age, whereas the intakes of most nutrients (from caloric beverages including milk) decreased, and energy and added sugar increased.SSBs, by definition, are beverages containing added sugar and/or sweetener and include regular soda, fruit drinks, sports drinks, energy drinks, presweetened ready-to-drink tea, and sweetened ready-to-drink coffee.SSB consumption begins during the preschool years and generally increases with age [22,23].In our analysis, we also found that intake of SSBs, including soda and fruit drinks, increased with age with a concomitant decrease in milk intake.These beverage consumption changes have been of public health concern due to their suspected linkage with health-related factors, such as excess energy intake, overweight and obesity, poor cardiometabolic health, and insulin resistance [24][25][26][27].Additionally, SSB consumption by young children and adolescents has been linked to reduced milk intake and negatively correlated with achieving intakes of vitamin A, vitamin C, and calcium [28][29][30].A number of studies have examined the impact of swapping various beverages with milk.In a randomized crossover study, a 3-wk intervention with milk replacing soda decreased systolic blood pressure, uric acid, and glycosphingolipids but did not affect lipoproteins in male adolescents [31].In a systematic review and meta-analysis of randomized control trials, substituting SSBs with flavored milk or noncaloric beverages significantly reduced body fat, with no change in BMI [32].In a European cohort of children, replacement of SSBs with water or milk was associated with reduced body fatness [33].In another study, although replacing habitual consumption of SSBs with milk did not affect the percent body fat, lean body mass was significantly increased [34].Isocaloric replacement of all caloric beverages (including soft drinks and SSBs) with milk at lunch and dinner was found to increase intake of protein and several key micronutrients in the present modeling analysis.Flavored milk is prepared from dairy milk and edible flavorings with or without caloric sweeteners.There is a perception by some that consumption of flavored milk will have a negative impact on the overall quality of children's diets due to added sugar, and some have called for banning flavored milk from school lunchrooms [35][36][37].However, flavored milk is a nutrient-rich beverage and has a nutrient profile similar to that of plain milk.Flavored milk is a rich source of calcium, protein, vitamin D, vitamin A, vitamin B 12 , potassium, phosphorus, riboflavin, and niacin.Although sweetened flavored milk has added sugar, in the United States, the contribution of added sugar from flavored milk is only ~3% of total energy [38] and is much less than that from SSBs [39].A review of scientific studies concluded that flavored milk consumption was not associated with higher added sugar intake [40].Flavored milk is a palatable beverage choice that helps children to meet their nutrient needs, and children who consumed flavored milk were found to consume more total milk and had higher intakes of nutrients, specifically calcium, phosphorus, magnesium, potassium, and vitamins A and D in multiple scientific studies [16,17,38,[40][41][42][43].Acceptability of flavored milk, especially chocolate milk, among children is much higher than that of plain milk, and studies suggest that children choose flavored over plain milk and consume ~30% more total milk [ [44]; M. Read, K. Henderson, M. Schwartz, unpublished results, 2011].Additionally, in a modeling study [45], removal of flavored milk from school meals resulted in decreased milk consumption and increased milk waste, and to make up for lost nutrients, 3-4 additional foods (including cheese, baked beans, vegetables, and fruit yogurt) were consumed with additional fat and calories at an increased dollar amount (~$50 per student per year).In the present study, flavored milk intake was highest among children aged 6-8 y and consistently decreased with age, whereas children aged 14-18 y consumed ~62% less flavored milk than younger children.Encouraging older adolescents to replace SSBs with flavored milk could benefit nutrient intake and lessen the gap between current dairy intake recommendations.
Milk substitutes, including almond milk, soy milk, rice milk, oat milk, and coconut milk, are nondairy beverages made from plant-based ingredients (such as rice, nuts/seeds, coconut, oats, peas, or blends of these ingredients).They are available in various flavors and options such as plain, vanilla, or chocolate, as well as organic, fortified or unfortified, and sweetened (with added sugar) or unsweetened.They have become popular over the past 2 decades; however, they are often a nutritionally inferior substitute to dairy milk with the exception of fortified soy, and scientists have advocated for nutritional standards for these beverages [46,47].In the present analysis, intake of these milk substitutes was very small (<5 g/d) and was not affected by age.
Dairy (including milk, cheese, and yogurt) is an important food group and an important component of a healthy diet and is included in the Healthy Dietary Patterns developed by USDA and released as part of Dietary Guidelines for Americans  [48].Despite these recommendations, a vast majority of Americans (>80%) consume less than the recommended intake of dairy [1].Although adequate dairy intake is associated with reduced risk of various noncommunicable chronic diseases [49], lower dairy intake among children and adolescents not only affects their growth and development but impacts their adult life.A study published in 2012 reported that childhood milk intake was positively associated with physical performance in old age [50].There is an increasing body of evidence to suggest that dairy consumption, including total dairy and fuller fat milk, may be beneficial to overall nutrient status and cardiometabolic health [32,51,52].Intake of dairy products including milk has been associated with reduced prevalence and risk of obesity among children [32,[52][53][54][55] and is beneficial for reducing the risk of a variety of chronic diseases in adulthood [51,56].However, beverage milk consumption has been consistently declining since 1975 [57].Additionally, due to the ongoing debate about the environmental impact from animal agriculture, some are advocating to remove or limit animal-sourced foods, including dairy [58,59].However, limiting relatively low-cost nutrient-dense foods, such as milk or substituting milk with other caloric beverages, such as soft drinks and SSBs could have potential unintended consequences, such as increased nutrient inadequacy and/or deficiencies [60][61][62].
Strengths of this study include the use of large nationally representative sample achieved through combining several sets of NHANES data releases and the same-person modeling used in this study.A major limitation of this study is the use of self-reported dietary intake data, which therefore are subject to reporting bias [63].Additionally, the dietary modeling approaches used here are focused on evaluating the maximum effect (assuming all subjects would comply with a modeled condition), may not reflect actual individual dietary behavior, and can only be used as an estimate of potential impact to nutrient intake as they cannot predict actual compliance in free-living individuals; however, such modeling offers a technique to test the potential nutritional impact of dietary guidance.Future directions should include modeling the replacement of all caloric beverages at mealtimes with commonly chosen plant-based milk alternatives to better understand the impact of mealtime choices and provide more informed guidance.
In conclusion, intake of SSBs, including soft drinks, other caloric beverages, and added sugar, as well as energy, increased with age whereas intake of milk decreased and consequently, the nutrient contribution of caloric beverages including milk also decreased.Additionally, the results of this study show that the addition of a serving of milk or isocaloric replacement of beverages with milk resulted in an increase in several micronutrients, including calcium, potassium, vitamin D, magnesium, and protein and no change or decrease in added sugar.The current findings provide additional evidence to support dietary recommendations for milk, and increased efforts appear to be needed to reverse the decrease in milk intake over time and as children age.

TABLE 1
Demographics of children by age groups, NHANES 2001-2018 Data presented as mean AE SEM.

TABLE 2
Mean intake of beverages among children of different age groups, sex-combined data NHANES 2001-2018 Data presented as least square mean AE SEM after adjusting for sex, ethnicity, and poverty-income ratio level.Regression coefficient represents the change across ages 6 to 18 y and indicates the change per 1 y of age.Abbreviation: SSB, sugar-sweetened beverage.

TABLE 3
Mean intake of nutrients from caloric beverages including milk among children of different age groups, sex-combined data NHANES 2001-2018 Data presented as least square mean AE SEM after adjusting for sex, ethnicity, and poverty-income ratio level.Regression coefficient represents the change across ages 6 to 18 y and indicates the change per 1 y of age.Abbreviations: DFE, dietary folate equivalent; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; RE, retinol equivalent; SFA, saturated fatty acid; tsp eq, teaspoon equivalent.

TABLE 4
Mean intake of nutrients among children of different age groups after addition of 1 cup eq of milk, sex-combined data NHANES 2001-2018 Data presented as least square mean AE SEM after adjusting for sex, ethnicity, and poverty-income ratio level.Abbreviations: DFE, dietary folate equivalent; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; RE, retinol equivalent; SFA, saturated fatty acid; tsp eq, teaspoon equivalent.

TABLE 5
Mean intake of nutrients among children of different age groups after isocaloric replacement of all caloric beverages during lunch and dinner with milk, sex-combined data NHANES 2001-2018 2020-2025 [1] and in MyPlate recommendations