Total beverage consumption and beverage choices among children ...

International Journal of Food Sciences and Nutrition, Volume 54, Number 4 (July 2003) 297 /307

Total beverage consumption and beverage choices among children and adolescents Richard A. Forshee and Maureen L. Storey Center for Food and Nutrition Policy, Virginia Polytechnic Institute and State University, 1101 King Street, Suite 611, Alexandria, VA 22314, USA

Patterns of beverage consumption among children and adolescents are related to age, race, and gender. The relationship between body mass index (BMI) and beverage consumption is unclear. In this paper, the total amount and the types of beverages consumed were analyzed according to age, race, and gender. Multivariate regression models were estimated for consumption of milk, juices, fruit drinks/ades, and carbonated soft drinks. Descriptive and multivariate regression analysis of children aged 6 /19 from the US Department of Agriculture’s Continuing Survey of Food Intake by Individuals 1994/96, 98 was performed. It was found that age, race, and gender play a significant role in the total amount, types, and relative proportions of beverages consumed by children and adolescents. Individuals in the first decile drink approximately 212.9 g beverages per day, whereas individuals in the tenth decile drink 2036.2 g. Boys drink more of most beverages than girls do. Older teens tend to drink more carbonated beverages, fruit drinks/ades, and citrus juice, but less fluid milk and non-citrus juice. White adolescent boys are heavy consumers of most beverages, including carbonated soft drinks, milk, and fruit drinks/ades. BMI is positively associated with consumption of diet carbonated beverages and negatively associated with consumption of citrus juice. BMI was not associated with consumption of milk, regular carbonated beverages, regular or diet fruit drinks/ades, or non-citrus juices. In conclusion, total beverage consumption and beverage choices are strongly related to age, race, and gender. BMI was only related to consumption of diet carbonated beverages and milk, and those relationships were weak.

Introduction Patterns of beverage and fluid consumption by children and adolescents can have a strong relationship with their overall nutrition status and are relevant to several contemporary

nutrition policy debates. Beverage and fluid consumption is influenced by many factors including age, race, gender, and physical activity, but we know very little about the

Correspondence to: Maureen L. Storey. e-mail: [email protected]

ISSN 0963-7486 printed/ISSN 1465-3478 online 03/010005-10 # 2003 Taylor & Francis Ltd DOI: 10.1080/09637480120092143

298

R. A. Forshee and M. L. Storey

consumption patterns of children and adolescents or about the demographic factors related to beverage choices. While media reports have stirred great controversy about beverage consumption among children and adolescents, scientists understand very little about total beverage consumption and patterns that may be influenced by age, gender, race/ethnicity, and body size (body mass index (BMI)). Most of the published literature on children’s beverage consumption focuses on hypothesized trade-offs, the influence of parental beverage choices, or energy intake contributed by beverages (Guenther, 1986; Johnson et al. , 1998; Harnack et al. , 1999; Troiano et al. , 2000; Fisher et al. , 2001). Birch and associates have found that adult role models are important in setting up behavioral characteristics that influence beverage choices (Fisher & Birch, 1999a,b; Fisher et al. , 2001; Lee et al. , 2001). Moreover, restricting access to palatable foods does not teach children to moderate consumption of those foods. We are not aware of any published studies examining the total beverage consumption and beverage choices made by children and adolescents. While commonsense tells us that older adolescents will drink more than younger children, there are no published data confirming that. Moreover, there are no published data examining the total beverage consumption or beverage choices that may be influenced by gender or race. Nor are there any published reports that overweight children and adolescents drink more, and hence consume more energy from beverages, than children and adolescents who are not overweight. Throughout this paper, beverage consumption will mean fluid milk, carbonated soft drinks, fruit drinks/ades, and citrus and noncitrus juices. Alcoholic beverages, coffee, and tea were excluded because few children and adolescents consume them. Liquids that might be construed as a meal, such as soups or meal replacements, were excluded from the analysis. This study has three goals: to examine total beverage consumption among children and adolescents, to examine their beverage choices, and to understand the

relationships between beverage consumption and choice as influenced by age, gender, race/ ethnicity, and BMI (i.e. body size). Data and methods The data in this paper are from the US Department of Agriculture’s (USDA) Continuing Survey of Food Intake by Individuals (CSFII) 1994/96, 98. Data used in this study were collected by the USDA and represent all non-institutionalized persons older than age 2 residing in the United States. The methods used to collect the data have been reported previously (US Department of Agriculture, 1998). Throughout this paper, beverage consumption data do not include water consumption, which is difficult to assess. Total non-water beverage consumption is calculated as the sum of fluid milk, fruit drinks/ ades, carbonated soft drinks, citrus juice, and non-citrus juice. Analysis of variance with post-hoc multiple comparisons was used to determine differences in mean beverage consumption as influenced by race/ethnicity. Multivariate regression analyses were used to model the consumption of individual beverage types as a function of gender, age, race, and BMI. The question is complicated because BMI increases as children get older, and there are gender and race differences in average BMIs. For example, African-American girls have a higher average BMI. Because of these interconnected relationships, multiple regression models were used to control for these factors. In these analyses, different categories of beverages were the dependent variables. The independent variables included BMI, age, race, and family income. A set of dummy variables was used to identify the race of the respondent, and separate models were estimated for boys and girls. The models were estimated in Stata 6 using the survey regression procedure (svyreg) to adjust for the complex sample design of CSFII. The numbers of children (age 6/11 years) and adolescents (age 12/19 years), gender, and race/ethnicity are presented in Table 1. Individuals of ‘other races’ are included in the totals of different categories but, because of the small numbers and heterogeneous mix

Beverages among children and adolescents

299

Table 1. Description of sample (unweighted)

Male (n /1687) Female (n /1624) Total (n /3311)

Non-Hispanic white

Non-Hispanic African-American

Hispanic

Other race

Total

1110 1019 2129

226 274 500

270 261 531

81 70 151

1687 1624 3311

of races and ethnicities included in the category, these data are unreliable with respect to predicting behaviors.

Results The data illustrated in Figure 1 show marked differences in total beverage consumption between boys and girls, and beverage choices across deciles of intake among individuals aged 6 /19 years. Individuals in the first decile drink approximately 212.9 g beverages per day, whereas individuals in the tenth decile drink 2036.2 g. The figure also shows that the mean consumption of all beverages generally increases across the deciles of total consumption. Some of this increase in total beverage consumption is a function of age

and gender. For example, the average age of the individuals in the first and tenth deciles is 12.1 and 15.1 years, respectively. Moreover, 65.1% of the individuals in the first decile are female, whereas only 19.5% in the tenth decile are female. It is not surprising that adolescents drink significantly more beverages than children do and that boys of all ages tend to drink more than girls do. Young boys (age 6 /11 years) drink more than young girls do: 776 and 692 g/day, respectively. Average intake of all beverages by adolescent (12 /19 years) boys and girls is 1196 and 799 g/day, respectively. There are differences in total beverage consumption among children and adolescents of different races/ethnicities. Data presented in Figures 2 and 3 show differences in total beverage consumption by white, African-

Figure 1. Total consumption and beverage choices of all children aged 6 /19 years.

300


Figure 2. Total beverage consumption of boys and girls aged 6 /11 years.

Figure 3. Total beverage consumption of boys and girls aged 12 /19 years.


American, and Hispanic children and adolescents. White boys aged 6 /11 years drink significantly more than African-American boys, but not more than Hispanics; young girls drink about the same amount regardless of race/ethnicity. Among adolescents, white boys drink significantly more than AfricanAmerican or Hispanic boys, and white girls drink more than African-American girls, but not Hispanic girls. Except for consumption of citrus and noncitrus juices, beverage choices and consumption of some beverages is linked to age, gender, and race/ethnicity. Milk is the predominant beverage of choice among young boys and girls of all races/ethnicities studied, followed by carbonated soft drinks, fruit drinks/ades, citrus juices, and non-citrus juices (Table 2). There are, however, some differences in level of consumption across race/ethnicity. Young white boys and girls drink significantly more milk than do African-American boys and girls. Hispanic girls drink significantly more milk than AfricanAmerican girls, but there was no difference in milk consumption between African-American and Hispanic boys. Among the younger children, race/ethnicity was linked to consumption of fruit drinks/ades but not carbonated soft drinks. African-American children drink significantly more fruit drink/ades than do white children, but not more than Hispanic children.

Except for citrus and non-citrus juices, differences in beverage choices among adolescents become more exaggerated, especially between genders and among races/ethnicities (Table 3). Milk consumption falls in adolescence, and carbonated soft drink and fruit drinks/ades consumption increases. White teenage boys and girls drink significantly more milk than African-American teenagers. White adolescent boys drink significantly more carbonated soft drinks than do African-American or Hispanic adolescent boys. White adolescent girls drink significantly more of this beverage than African-American girls, but not more than Hispanic adolescent girls do. African-American adolescent girls drink about 60% less carbonated soft drinks, but 60% more fruit drinks/ades than white and Hispanic adolescent girls. In general, African-American females drink less milk than others do. Young African-American girls drink 209 g/day, which drops one-half during adolescence. On average, African-American adolescent girls drink 103 g (3.4 oz) of milk per day. Because average milk consumption among African-American adolescent girls was so low, we examined the frequency data for patterns of consumption. We found an apparent ‘all-or-none’ phenomenon in this group of girls. About one-half (48.8%) of African-American adolescent girls reported drinking no milk on the days of the survey, and 89% of this group reported drinking less than one cup. This contrasts with their

Table 2. Beverage choices (g/day) among boys and girls aged 6 /11 years Milk

Soda

Fruit ade

Citrus juice

Non-citrus juice

Total

Boys (n /748) White African-American Hispanic Other race Total

344a 255b 335a,b 308a,b 327

215a 175a 201a 201a 206

135a 201b 150a,b 176a,b 150

61a 42a 54a 41a 56

37a 38a 38a 34a 37

838a 742b 799a,b 787a,b 816

Girls (n /712) White African-American Hispanic Other race Total

284a 209b 314a 282a,b 276

205a 160a 160a 131a 188

126a 179b 119a,b 68a 131

51a 58a 51a 70a 53

49a 33a 34a 24a 44

748a 661a 718a 592a 724

a,b

301

Means within gender in the same column bearing different superscripts differ significantly (p B/0.05).

302


Table 3. Beverage choices (g/day) among boys and girls aged 12 /19 years Milk

Soda

Fruit ade

Citrus juice

Non-citrus juice

Total

Boys (n /941) White African-American Hispanic Other race Total

329a 189b 284a,b 220a,b 299

671a 407b 454b 352b 593

183a 321b 181a 69a 195

81a 86a 85a 56a 81

29a 24a 21a 50a 28

1406a 1132b 1106b 860b 1305

Girls (n /909) White African-American Hispanic Other race Total

197a 103b 194a 157a,b 178

431a 263b 372a,c 273b,c 387

122a 205b 120a 61a 134

64a 67a 89a 65a 68

36a 26a 16a 51a 33

a,b

954a 741b 858a,b 728b 893

Means within gender in the same column bearing different superscripts differ significantly (p B/0.05).

childhood choices, showing that only 13% of young African-American girls drank no milk. In comparison, about one-third of white and Hispanic adolescent girls report no milk consumption on the days of the survey. We also examined the distribution of carbonated soft drink and fruit drink/ades consumption among children and adolescents. Consumption of carbonated soft drinks was the same among boys and girls of different races, with an average of 206 and 188 g/day (6.6 and 6.1 oz/day), respectively. About 36% of young white and Hispanic boys and girls report drinking no carbonated soft drinks, whereas nearly one-half of African-American boys and girls did not drink them. Among the teenagers, between 12 and 14% of white and Hispanic boys drank no carbonated soft drinks, but nearly 28% of AfricanAmerican adolescent boys reported drinking none. Approximately 18% of white adolescent girls reported no carbonated soft drink consumption, but 29% of African-American and 25% of Hispanic girls reported drinking none. As an internal validity check, we compared our findings on the frequency of beverage consumption with the USDA CSFII Table Sets. As expected, our findings are consistent with those reported by the USDA. The USDA reports that 47.1% of boys 6 /11 years of age and 44.8% of girls 6/11 years of age consume some carbonated soft drinks. In the

12 /19 year age range, 69.2% of boys and 62.2% of girls consume some carbonated soft drinks. The remaining children did not report any consumption of carbonated soft drinks (US Department of Agriculture, 1999). The heaviest consumers of carbonated soft drinks are white adolescent boys who consume about 671 g (1.8 /12-oz cans) per day. The boys in the 90th percentile of consumption drink 1364 g or about 3.7 /12-oz cans per day. Regression analysis In the regression analysis, we examined multiple factors that may influence total beverage consumption and beverage choices, as shown in Table 4. The race/ethnicity differences in beverage consumption discussed in the previous section are confirmed after controlling for BMI, age, and family income. In general, AfricanAmerican children drink less fluid milk and carbonated soft drinks but more fruit drinks/ ades than white children. The race/ethnicity differences in beverage consumption remained even after the inclusion of family income in the model. That strongly suggests that the race/ethnicity differences in beverage consumption are not a function of differences in family income. Family income, as measured by income as a percent of the poverty line, has very little association with beverage consumption. Boys and girls in higher income families consume

Table 4. Unstandardized regression coefficients and t -ratios of beverages consumed by boys and girls

Variable Boys Body mass index Age Income AfricanAmerican Hispanic Other race Constant

Diet carbonated soft drinks

Regular fruit drinks and ades

Diet fruit drinks and ades

4.8 (1.29)

1.6 (1.99)*

/1.9 (/0.81)

0.5 (0.41)

/1.3 (/1.54)

0.7 (1.03)

4.2 (1.05)

/7.1 (/2.49)* 0.2 (1.33) /97.0 (/2.56)*

51.3 (10.52)* 0.3 (1.26) /148.1 (/3.74)*

/0.7 (/0.67) 0.1 (2.33)* /22.7 (/3.78)*

3.5 (1.18) /0.1 (1.29) 120.3 (2.47)*

3.1 (2.13)* 0.1 (1.67) /24.3 (/3.59)*

3.8 (2.65)* 0.2 (2.84)* 4.6 (0.35)

/1.9 (/2.48)* 0.1 (3.21)* 5.4 (0.78)

52.1 (8.32)* 0.7 (3.03)* /160.8 (/2.97)*

/33.2 (/0.97) /82.0 (/1.77) 378.7 (5.70)*

/107.0 (/2.31)* /165.3 (/2.82)* /352.8 ( /4.55)*

/18.9 (/3.17)* /21.2 (/3.18)* /11.0 ( /0.72)

16.9 (0.88) /25.2 (/0.87) 152.9 (3.76)*

/8.8 (/0.92) /18.3 (/1.63) /28.1 ( /1.37)

14.7 (1.42) /19.3 (/1.05) 11.0 (0.43)

/3.5 (/0.54) 15.0 (1.50) 23.5 (1.63)

/143.7 (/2.48)* /319.4 (/3.45)* 184.6 (2.11)*

/4.2 (/3.69)*

0.4 (0.21)

2.8 (2.42)*

/1.4 (/1.42)

0.7 (1.59)

/0.5 (/0.99)

/0.5 (/0.77)

/2.7 (/0.88)

/11.5 (/5.75)* 0.0 (/0.58) /81.7 (/4.49)*

23.2 (7.99)* /0.1 (/0.66) /92.0 (/3.22)*

1.6 (2.14)* 0.1 (3.38)* /28.1 (/3.07)*

1.0 (0.48) 0.0 (/0.33) 67.7 (3.76)*

0.5 (0.68) 0.0 (1.06) 7.4 (0.90)

1.5 (1.69) 0.1 (2.00)* 8.7 (0.63)

/1.3 (/1.29) 0.0 (1.47) /7.1 (/0.81)

15.0 (4.13)* 0.1 (1.27) /124.4 (/3.43)*

0.4 (0.02) /27.7 (/0.80) 472.4 (13.10)*

/20.1 (/0.53) /133.9 (/3.35)* 13.2 (0.25)

/22.5 (/1.90) 13.3 (0.40) /62.4 (/2.46)*

/8.2 (/0.61) /63.2 (/2.94)* 129.8 (4.24)*

/2.0 (/0.40) 1.4 (0.09) /12.8 (/1.16)

21.6 (2.07)* 10.3 (0.57) 30.2 (1.95)

/16.0 (/2.00)* 5.5 (0.39) 57.9 (3.65)*

/49.3 (/1.27) /198.0 (/2.74)* 627.8 (10.36)*

0.0 (/0.01)

*Indicates the coefficient is statistically significant at the p B/0.05 level.

Citrus juice

Non-citrus juice

Total beverages


Girls Body mass index Age Income AfricanAmerican Hispanic Other race Constant

Regular carbonated soft drinks

Milk

303

304


slightly more diet carbonated soft drinks. For boys, family income also has a slight positive association with consumption of citrus juice and non-citrus juice. All of the family income associations are very weak. Age has a negative association with consumption of fluid milk and a positive association with consumption of carbonated soft drinks. A year of age is associated with consumption of 7 g fewer of milk for boys and 12 g fewer of milk for girls. A year of age is associated with consumption of 51 g more (1.6 oz) of regular carbonated soft drink for boys and 23 g more (0.7 oz) of regular carbonated soft drink for girls. Consumption of diet carbonated soft drinks also increases slightly for older girls (1.6 g or about onetwentieth of an ounce for each year of age). Older boys drink slightly more diet fruit drinks/ades and citrus juice but slightly less non-citrus juice. BMI has little or no association with beverage consumption in these models. In other words, heavier individuals do not drink more beverages than lighter individuals. After controlling for age, race/ethnicity, and family income, BMI has no statistically significant relationship with any of the beverages for boys. For girls, BMI has a slight negative association with milk consumption and a slight positive association with consumption of diet carbonated soft drinks. Each point of BMI is associated with 4 g fewer milk consumption and 3 g more diet carbonated soft drink consumption. These predicted associations amount to about onetenth of an ounce per BMI point. Preliminary research that we reported showed that children above the 85th percentile of BMI consumed about 2 oz more of total carbonated soft drinks than did those below the 85th percentile. That preliminary research had several limitations that have been improved in this version. Most importantly, at the time that research was presented, age-specific BMI guidelines were not available, so the 85th percentile for age ranges was used. Since age, BMI, and total carbonated soft drink consumption are all positively correlated, a disproportionate number of the children above the 85th percentile were older children, some of

whom would not be above the 85th percentile using the new, age-specific guidelines. A second important limitation of the earlier research is that it did not disaggregate total carbonated soft drink consumption into regular and diet carbonated soft drink consumption. The Centers for Disease Control and Prevention have established age and genderspecific growth charts that specify the 85th percentile of BMI as the threshold for being at risk of overweight (Kuczmarski et al ., 2000). We therefore estimated separate models in which the continuous variable for BMI was transformed to a binary variable indicating whether the individual was below the 85th percentile of BMI (non-overweight) or at or above it for their age and gender. This further tested whether children with a high BMI consumed more beverages. The results were nearly identical to the results of the continuous model. Controlling for the other factors in the model, girls with BMI above the 85th percentile consumed 24 g fewer of milk. The 85th percentile BMI variable was not statistically significant among girls for any of the other beverages. Boys in the 85th percentile of BMI did not consume more milk, regular carbonated soft drinks, fruit drinks/ades, or citrus juices than non-overweight boys did. Boys above the 85th percentile consumed 12 g more (0.4 oz) of diet carbonated soft drinks and 12 g more noncitrus juices. Discussion The present study investigated various influences on total consumption and choices of beverages. As expected, age and gender influenced total beverage consumption */ older individuals drink more than younger individuals, and boys drink more of most beverages than girls. Differences in total beverage consumption and choices among individuals of different races/ethnicities have not been reported previously and were unexpected. Overall, this study shows that, with one exception, white children and adolescents drink more beverages than their African-American counterparts. The reasons for this are unknown and should be explored


further, especially with regard to proper hydration for everyday activity and during strenuous physical activity. Beverage choices were notably different depending on race/ethnicity. While young boys and girls of all races/ethnicities drink more milk than any other beverage, AfricanAmerican children drink less milk than white or Hispanic children. Young white boys and girls consume about 1.4 and 1.2 servings of milk per day, whereas young African-American boys and girls drink 1.0 and 0.8 servings per day, respectively. Young Hispanic girls might be considered the most avid milk consumers, with few of them eschewing milk entirely */ 7.4% reported drinking no milk. On the other hand, 13 /14% of young African-American children report drinking no milk. It was surprising that African-American adolescent girls drank so few beverages as a whole, and so little milk in particular. This is a concern not only for future bone health, but also because calcium consumption may be important in reducing the risk of hypertension */ a serious condition that is more prevalent in African-Americans. Proper hydration is another concern unless AfricanAmerican adolescent girls are consuming enough water to compensate for the lack of other fluid intake. Among the beverage alternatives, carbonated soft drinks are the second or third choice of young children, depending on race/ ethnicity. On average, young boys and girls drink 206 g (6.6 oz) and 188 g (6.2 oz) of carbonated soft drinks per day, which represents about 4% of total daily energy intake, or about 77 and 72 kcal, respectively. This amount of caloric intake might easily be balanced with even a modest level of physical activity during recess, physical education classes at school, or after-school activity programs. During adolescence, milk consumption falls, becoming the second most consumed beverage (on a gram basis) following carbonated soft drinks. Older teens tend to drink more carbonated beverages, fruit drinks/ ades, and citrus juice, and less fluid milk and non-citrus juice. White adolescent boys, in particular, are heavy consumers of most

305

beverages, including carbonated soft drinks, milk, and fruit drinks/ades; on average, they drink the equivalent of about 1.8 /370 g cans of carbonated soft drinks per day, compared with 1.0 /370 g cans and 1.2 / 370 g cans for African-American and Hispanic adolescent boys, respectively. Beverage consumption data are similar between the two largest, federally funded dietary surveys */ NHANES III and CSFII. Troiano et al . (2000) examined data from the NHANES survey and reported that beverages are a significant source of total energy for 6- to 11-year-old children and for 12- to 19-year-old adolescents, contributing 20/ 22% of energy intake in these age groups. These authors noted that milk contributed 50% or more of the energy from beverages for children younger than 12 years of age. Among adolescents, carbonated soft drinks provided 8% of the total energy intake (Troiano et al. , 2000). Moreover, an erratum of research published by Cavadini et al. (2001) showed that adolescent boys and girls drink about 1.4 /370 g cans and 0.9 /370 g cans of soft drinks per day, respectively. Our analysis of the CSFII survey data is nearly identical with these findings; that is, on average, adolescent boys and girls consume about 1.5 /370 g cans and 1.1 /370 g cans of carbonated soft drinks per day, respectively, which is about 8% of the total daily energy intake. Energy intake must be balanced with energy expenditure if individuals, whether they are growing or not, are to avoid excess weight gain. Recently, Johnson & Frary (2001) reviewed the few studies that directly examine the role of added sugars and carbonated soft drinks in rising rates of overweight among children and adolescents. Harnack et al. (1999) found greater energy intake among 6- to 12-year-old children who drank 9 oz or more of soft drinks than those who did not drink them. The authors did not report BMI or activity levels of these children, and included only carbonated soft drinks and total energy in the analysis. In an 18-month longitudinal study, Ludwig et al. (2001) noted that middle-school children who consumed more carbonated soft drinks were more likely to become at risk of overweight

306


*/ at or above the age-specific and genderspecific 85th percentile of BMI. The authors, however, did not assess the beverage consumption habits of the children whose BMI changed to below the 85th percentile, and hence would no longer be classified as at risk of overweight. The multivariate regression models reported here show no association between age-specific and gender-specific BMI and regular carbonated soft drink consumption, but there is a positive association between BMI and diet carbonated soft drinks. It is not clear how much energy should be derived from beverages for children and adolescents for proper hydration and nutrient contributions to the diet. To provide perspective, two servings of reduced fat (2%) milk plus one serving of orange juice provide 328 kcal energy, which is approximately 12% of total energy for adolescent boys. Adding the caloric contribution of 1.5 /370 g servings (210 kcal) of a regular carbonated soft drink brings the total consumption to 538 kcal */ approximately 19% of total energy intake. The energy contribution by these beverages appears to be reasonable for an individual engaged in even a modest level of physical activity. As we previously testified, the US Dietary Guidelines for Americans Committee, age, gender, and race/ethnicity are non-modifiable and strong predictors of BMI among children and adolescents. This present study shows that these non-modifiable factors are also strong predictors of total beverage

consumption and choices. Controlling for these factors and BMI are therefore critical to understanding the relationship with beverage consumption. This analysis refutes widespread speculation that carbonated soft drinks are responsible for the increase in overweight among children and adolescents. While some individuals may consume large amounts of carbonated soft drinks and/or fruit drinks/ades, it does not appear to be the typical consumption pattern. Moreover, the reasons for consuming larger than average amounts of these beverages should be explored, especially among children and adolescents who are engaged in strenuous physical activity. Careful monitoring of children’s beverage intake is nevertheless warranted because caloric contributions must be balanced with energy expenditure. In addition, consumption of milk should be encouraged through continuation of strong promotional efforts, but most importantly through parental role models. But draconian policies aimed at curtailing consumption of carbonated soft drinks and fruit drinks/ades are impermanent solutions that will fail to coerce children and adolescents, in particular, to reject soft drinks.

Acknowledgements */An earlier version of this research that was presented at the Federation of American Societies for Experimental Biology, San Diego, CA, April, 2000. The research presented in this manuscript was supported by a grant from the National Soft Drink Association.

References Cavadini C, Siega-Riz AM & Popkin BM (2001): U.S. adolescent food intake trends from 1965 to 1996. West. J. Med. 175, 235. Fisher JO & Birch LL (1999): Restricting access to palatable foods affects children’s behavioral response, food selection, and intake. Am. J. Clin. Nutr. 69, 1264 /1272. Fisher JO & Birch LL (1999): Restricting access to foods and children’s eating. Appetite 32, 405 /419. Fisher J, Mitchell D, Smicklas-Wright H & Birch LL (2001): Maternal milk consumption predicts the tradeoff between milk and soft drinks in young girls’ diets. J. Nutr. 131, 246 /250.

Guenther PM (1986): Beverages in the diets of American teenagers. J. Am. Diet Assoc. 86, 493 /499. Harnack L, Stang J & Story M (1999): Soft drink consumption among US children and adolescents: nutritional consequences. J. Am. Diet Assoc. 99, 436 /441. Johnson RK & Frary C (2001): Choose beverages and foods to moderate your intake of sugars: the 2000 Dietary Guidelines for Americans */ what’s all the fuss about? J. Nutr. 131, 2766S /2771S. Johnson RK, Panely C & Wang MQ (1998): The association between noon-time beverage consumption and the diet quality of school-aged children. J. Child Nutr. Manage. 2, 95 /100.

Beverages among children and adolescents Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, Mei Z, Curtin LR, Roche AF, Johnson CL (2000): CDC Growth Charts: United States, Advance data from vital and health statistics, No. 314. Hyattsville, MD: National Center for Health Statistics [http://www.cdc.gov/nchs/data/ad/ad314.pdf], accessed 11 December 2001. Lee Y, Mitchell DC, Smiciklas-Wright H & Birch LL (2001): Diet quality, nutrient intake, weight status, and feeding environments of girls meeting or exceeding recommendations for total dietary fat of the American Academy of Pediatrics. Pediatrics 107, E95. Ludwig DS, Peterson KE & Gortmaker SL (2001): Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 357, 505 /508.

307

Troiano RP, Briefel RR, Carroll MD & Bialostosky K (2000): Energy and fat intakes of children and adolescents in the United States: data from the national health and nutrition examination surveys. Am. J. Clin. Nutr. 72, 1343S /1353S. US Department of Agriculture (1998): 1994 /96 Continuing Survey of Food Intakes by Individuals and 1994 /96 Diet and Health Knowledge Survey and related materials (CDROM). Beltsville, MD: Agricultural Research Service. US Department of Agriculture (1999): Food and Nutrient Intakes by Children 1994 /96, 1998 , Table 15B. ARS Food Surveys Research Group, Beltsville, MD: Agricultural Research Service [http://www.barc.usda.gov/bhnrc/foodsurvey/home.htm] (‘products’ page), accessed 11 December 2001.