R

ABJ

ADNdccgcssirado5caCstJ

Maccate(db

JWpCKC

Ci

A

5

RESEARCH

esearch and Professional Briefs

cceptability of Soymilk as a Calcium-Richeverage in Elementary School Children

ENNIFER K. REILLY, RD; AMY J. LANOU, PhD; NEAL D. BARNARD, MD; KIM SEIDL, MS, RD; AMBER A. GREEN, RD

itreccmtdi

topdslisaf

MSFfBaDBDf

um0ata2t

nbtlbA1

BSTRACTairy products are the main source of calcium in theational School Lunch Program. Children who avoidairy products because of lactose intolerance or otheroncerns can benefit from the availability of another cal-ium-rich beverage in the National School Lunch Pro-ram. This study assessed whether elementary schoolhildren would select and consume calcium-fortifiedoymilk in school cafeterias and whether the addition ofoymilk would increase the percentage of students select-ng a calcium-rich beverage at lunch. The ratio of satu-ated fat to calcium consumed from calcium-rich bever-ges with and without the availability of soymilk was alsoetermined. After 4 weeks of soymilk availability, 22.3%f students chose soymilk and consumed an average of8% of each carton, whereas 77.6% chose dairy milk andonsumed an average of 52.6%. Total calcium-rich bever-ge selection with lunch increased from 79% to 83.1%.alcium consumption per gram of saturated fat con-umed from calcium-rich beverages also rose from 194 mgo 237 mg with the addition of soymilk.Am Diet Assoc. 2006;106:590-593.

any US children are lactose intolerant or allergicto milk, or choose to avoid milk for reasons such astaste preferences, religious or ethical consider-

tions, or health concerns (1-4). According to the Ameri-an Academy of Family Physicians, the majority of Afri-an Americans, Asian Americans, Hispanic Americans,nd Native Americans are lactose intolerant, and symp-oms typically develop gradually during childhood andarly adulthood (5). The National School Lunch ProgramNSLP), which serves low-cost and free meals to US chil-ren, currently does not offer a reimbursable alternativeeverage to dairy milk without a doctor’s request (6). This

. K. Reilly is managing director, The Cancer Project,ashington, DC. A. J. Lanou is assistant professor, De-

artment of Health and Fitness, University of Northarolina at Asheville. N. D. Barnard is president and. Seidl and A. A. Green are staff dietitians, Physiciansommittee for Responsible Medicine, Washington, DC.Address correspondence to Jennifer K. Reilly, RD, Theancer Project, 5100 Wisconsin Ave, Suite 400, Wash-

ngton, DC 20016. E-mail: jreilly@cancerproject.orgCopyright © 2006 by the American Dietetic

ssociation.0002-8223/06/10604-0011$32.00/0

adoi: 10.1016/j.jada.2006.01.010

90 Journal of the AMERICAN DIETETIC ASSOCIATION

s noteworthy because the estimated mean calcium in-ake of children ages 6 to 11 years is well below theecommended amount (7). Enriched soymilk contains anquivalent amount of calcium, vitamin D, and vitamin Aompared with dairy milk, but contains no lactose orholesterol. It is low in saturated fat compared with dairyilk, which, in a National Cancer Institute study, was

he greatest single contributor to saturated fat in chil-ren’s diets (8). Calcium absorption from soymilk is sim-lar to that of dairy milk (9).

Before offering soymilk as a reimbursable beverage inhe NSLP, it must be deemed acceptable to children. Tour knowledge, this acceptability has not been testedreviously. This study, therefore, assessed whether chil-ren would select and consume soymilk when offered iningle-serving containers alongside dairy milk in schoolunch cafeterias during a 4-week period at three schoolsn southern Florida. We also evaluated the percentage oftudents selecting a calcium-rich beverage with lunchnd the ratio of saturated fat to calcium consumptionrom calcium-rich beverages with the addition of soymilk.

ETHODStudy Sampleor a 4-week study period, cartons of chilled, calcium-

ortified vanilla and chocolate soymilk (Silk, White Wave,oulder, CO) were available in the cafeteria lunch linest Dillard Elementary School (Broward County Schoolistrict), North Beach Elementary School, and Ruth K.road Bay Harbor Elementary School (both in Miami-ade School District) (n�3), where students range in age

rom 5 through 12 years.Soymilk cartons contained slightly (0.25 oz) more prod-

ct, compared to the 8.0 oz in a standard pint of dairyilk. Soymilk used in this study contained 3.5 g fat and

.5 g saturated fat per 8.25-oz serving. Dairy milk offeredt these schools contained the following quantities ofotal and saturated fat per 8-oz serving: skim milk, 0.5 gnd 0.3 g; chocolate milk, 1.5 g and 1.0 g; low-fat milk,.5 g and 1.5 g; and whole milk, 8.0 g and 5.0 g, respec-ively.

These three schools were chosen because they wereear one another, included students from differing racialackgrounds, and were interested in offering soymilk. Aotal of 1,850 to 2,025 free, reduced-price, and full-priceunches are served daily at these three schools to a com-ined student population that is 36.4% Hispanic, 34%frican American, 26.1% white, 2.3% multiracial, and.2% Asian (10).Because the study involved human subjects, it was

pproved by an Institutional Review Board (Independent

© 2006 by the American Dietetic Association

Rodcictstssbbpsl

SPdpclc

ssstt

dwssBdls

p

depcswcds

RDdra

ATfpsasdlSrsAt

AMdsa

eview Consulting, Inc [www.irb-irc.net]) prior to studynset. An adequate level of acceptability of soymilk wasefined a priori as 10% of all children purchasing lunchhoosing to purchase soymilk after 4 weeks of availabil-ty, with the average percentage of each soymilk cartononsumed (as opposed to discarded) similar to or greaterhan that of dairy milk. Soymilk acceptability at 10%election was determined based on market data showinghat soymilk represents 6% of all milk varieties sold inouthern Florida, combined with the fact that elementarychool children are encouraged to take a calcium-richeverage. The study period was set at 4 weeks’ durationased on the likelihood that new product novelty wouldersist for no more than approximately 2 weeks. Mea-ures of flavored dairy milk acceptability have been pub-ished previously, based on 2-day mean intakes (11).

tudy Designrior to introducing soymilk in the three schools, baselineairy-milk consumption was measured. Sales of eachroduct were tabulated. Milk cartons were collected ashildren returned their finished lunch trays during allunch periods for 2 consecutive days. Cartons wereounted and leftover milk was weighed.

On the following day, as they entered the cafeteria,tudents had a chance to sample the vanilla and chocolateoymilk that was to be offered in the lunch line. This gavetudents the opportunity to familiarize themselves withhe taste of soymilk and decide whether they would likeo select a whole carton starting the next day.

For the 2 days after soymilk sampling, soymilk andairy-milk measurements (selection and consumption)ere taken in the same way as baseline dairy-milk mea-

urements. For the next 4 weeks, vanilla and chocolateoymilk was placed in the cooler with the dairy milk.everages were rotated and arranged differently on aaily basis so that product position would not affect se-ection. At the end of 4 weeks, total soy and dairy-milkelection and consumption were measured as before.Broward County and Miami-Dade foodservice offices

Table 1. Average dairy milk and soymilk carton selection and conssoymilk introduction, and after 4 weeks of soymilk availability

Mean Carton Selection (%)

Prior tosoymilkintroduction

Initial 2 daysof soymilkintroduction

4afin

Chocolate dairy milk 79.6 36.0 61Whole dairy milk 17.9 9.6 11Lowfat dairy milk 2.5 1.2 5Skim dairy milk NAa NA 0Total 100 46.8 77

Chocolate soymilk NA 40.7 14Vanilla soymilk NA 12.4 7Total 53.1 22

aNA�not applicable.

rovided total student lunch sales on data-collection t

ays. Package labeling provided milk and soymilk nutri-nt contents. Investigators calculated calcium consumeder gram of saturated fat based on the weighted averageonsumption of each type of milk with correspondingaturated fat levels on study days. Mean calculationsere performed in a Microsoft Excel spreadsheet (Mi-

rosoft Corp, Redmond, WA). Data are presented withescriptive statistics only because the limited number ofchools precluded any testing for statistical significance.

ESULTSuring the preintroduction phase, 79% of students choseairy milk with lunch, whereas 83.1% chose a calcium-ich beverage (dairy or soy) with lunch after soymilk wasvailable for 4 weeks.

cceptability: Selectionable 1 shows average dairy-milk and soymilk selection

rom the lunch line as a percentage of total milk takenrior to introducing soymilk, during the first 2 days ofoymilk introduction, and after 4 weeks. Skim milk wasvailable sporadically at the three schools throughout thetudy period, but was unavailable at baseline due to pooremand; only 10 students chose skim milk on data-col-ection days—all were at North Beach Elementarychool. Chocolate dairy milk and chocolate soymilk rep-esented a total of 79.6%, 76.7%, and 75.6% of cartonselected at the three data-collection periods, respectively.fter 4 weeks, total soymilk selection included 22.3% of

he milk taken from the lunch line.

cceptability: Consumptionean consumption data for all types of milk at the three

ata points are presented in Table 1. Of the dairy milkelected, consumption differed by flavor, whereas vanilland chocolate soymilk consumption levels were similar

on at three elementary schools prior to introducing soymilk, during

Mean Carton Consumption (%)

soymilkction

Prior tosoymilkintroduction

Initial 2 daysof soymilkintroduction

4 weeksafter soymilkintroduction

61.5 62.2 53.848.0 50.4 47.634.0 39.4 33.2NA NA 39.258.8 59.4 52.6

NA 53.6 57.6NA 54.2 58.9

53.8 58.0

umpti

weekter strodu

.2

.1

.0

.3

.6

.4

.9

.3

hroughout the study period.

April 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION 591

CTgda

DIsWtpar

1spldchpda

saistdmspnw

cawtp

rsFwswlcssaaobccmp

CBtdtssnsspcdcsts

atf

TwtaBa

R

5

alcium-to-Saturated Fat Consumption Ratioable 2 shows the quantity of calcium consumed perram of saturated fat from dairy milk prior to the intro-uction of soymilk, and also after 4 weeks of soymilkvailability at each school.

ISCUSSIONn a 4-week study conducted at three elementary schools,oymilk proved to be an acceptable beverage to children.hen soymilk was available, more children took some

ype of calcium-rich beverage with their lunch, as com-ared to when only dairy milk was offered. Soymilk alsoppeared to have a favorable effect on the ratio of satu-ated fat to calcium from calcium-rich beverages.

Offering soymilk to children may be abeneficial way to lower their intake

of saturated fat.

In a previous study of soy acceptability, girls aged 8 to4 years adhered to an intervention that includedoymilk and soy nuts in the diet (12). In adults, soymilkroved more effective at lowering blood pressure andow-density lipoprotein cholesterol and increasing high-ensity lipoprotein cholesterol as compared with nonfatow’s milk (13,14). A recent study suggested that house-old availability of whole milk and other high-fat dairyroducts can contribute to overweight in Hispanic chil-ren (15). It is not clear that a change to soymilk wouldlter this finding.Strengths of this study include ethnic diversity of the

tudent population, an adequate time period for testing,nd objective measurements. Limitations of this studynclude uncertain generalizability, given that the threechools studied were all in southern Florida and in two ofhe largest school districts in the United States. Also, theifference in packaging of the soymilk and dairy milkay have affected selection. The soymilk was packaged in

hiny, prisma-pack containers, whereas dairy milk wasackaged in standard gable-top cartons. Investigators didot track whether the children had had milk or soymilk

Table 2. Calcium consumption per gram of saturated fat fromcalcium-rich beverages (dairy and soy) prior to the introduction ofsoymilk and after 4 weeks of soymilk availability at three elemen-tary schools

Calcium Consumptiona

Baseline dairy 4 weeks

Bay Harbor Elementary School 202 226North Beach Elementary School 183 214Dillard Elementary School 183 250

Average 194 237

aMilligrams per gram saturated fat.

ith breakfast or had other dairy products, such as

92 April 2006 Volume 106 Number 4

heese or ice cream, with lunch, nor did they measurembient temperature or beverage temperature, all ofhich may have influenced beverage choice. In addition,

he hot weather of southern Florida during the studyeriod may have affected consumption.Although the prisma-pack containers did not require

efrigeration, storage space and available refrigeratorpace to chill the drinks prior to lunchtime were limited.reezer space was often used for chilling the soymilk,hich may have affected consumption and acceptability if

ome of the soymilk froze. Delivery of soymilk from thearehouse to the schools also presented an initial chal-

enge because it was not a product the food distributionompany normally transported. In addition, foodservicetaff occasionally neglected to stock the milk coolers withoymilk, which meant that not all of the lunch periods onll the days in the 4-week study period had soymilkvailable. Also, in one school, the local dairy distributorsccasionally removed the soymilk from the lunch lineecause of their impression that the dairy-milk cratesould not be used to hold a competitive product. On data-ollection days, however, researchers ensured that dairyilk and soymilk were available throughout the lunch

eriods.

ONCLUSIONSy offering an additional type of calcium-rich beverage

hat was dairy-free, lactose-free, and acceptable to chil-ren, more children took a calcium-rich beverage withheir lunch and consumed a greater ratio of calcium toaturated fat from calcium-rich beverages. Enrichedoymilk may, therefore, be a useful option to meet theutrition needs of children served by the NSLP. Thistudy may be useful for dietetics professionals working inchool foodservices and with children who avoid dairyroducts, as it demonstrates calcium-fortified vanilla andhocolate soymilk as useful alternatives in meeting chil-ren’s calcium needs. Furthermore, offering soymilk tohildren may be a beneficial way to lower their intake ofaturated fat. In addition, serving calcium-rich beverageshat are shelf-stable may help foodservice directors solvetorage space issues.Further research may determine soymilk acceptability

mong different ethnic groups at schools participating inhe NSLP and assess the effects of packaging and otheractors in acceptability among schoolchildren.

his research was funded by The Cancer Project. Soymilkas provided by White Wave (Boulder, CO). The authors

hank Broward County and Miami-Dade School Districtsnd the foodservice directors at Dillard, North Beach, anday Harbor Elementary Schools for their assistance indministering the study.

eferences1. Gremse DA, Greer AS, Vacik J, DiPalma JA. Abdom-

inal pain associated with lactose ingestion in childrenwith lactose intolerance. Clin Pediatr (Phila). 2003;42:341-345.

2. DiPalma AM, DiPalma JA. Recurrent abdominalpain and lactose maldigestion in school-aged chil-

dren. Gastroenterol Nurs. 1997;20:180-183.

1

1

1

1

1

1

3. Goulding A, Rockell JE, Black RE, Grant AM, JonesIE, Williams SM. Children who avoid drinking cow’smilk are at increased risk for prepubertal bone frac-tures. J Am Diet Assoc. 2004;104:250-253.

4. American Dietetic Association, Dietitians of Canada.Position of the American Dietetic Association andDietitians of Canada: Vegetarian diets. J Am DietAssoc. 2003;103:748-765.

5. Swagerty DL, Walling AD, Klein RM. Lactose intol-erance. Am Fam Physician. 2002;65:1845-1850, 1855-1856.

6. US Department of Agriculture, Food and NutritionServices. National School Lunch Program. Alexan-dria, VA: US Department of Agriculture; 2003. Avail-able at: http://www.fns.usda.gov/cnd/Lunch/default.htm. Accessed December 3, 2004.

7. Continuing Survey of Food Intakes by Individuals(CSFII), Food and Nutrient Intakes by Children1994-96, 1998. Available at: http://www.ars.usda.gov/SP2UserFiles/Place/12355000/pdf/scs_all.PDF. Ac-cessed February 15, 2006.

8. Subar AF, Krebs-Smith SM, Cook A, Kahle LL. Di-etary sources of nutrients among US children, 1989-1991. Pediatrics. 1998;102:913-923.

9. Weaver CM, Plawecki KL. Dietary calcium: Ade-quacy of a vegetarian diet. Am J Clin Nutr. 1994;59(suppl 5):S1238-S1241.

0. Florida Department of Education’s School Advisory

Council Report and Florida School Indicators Report.Data current as of the 2002-2003 school year.Available at: www.greatschools.net. Accessed No-vember 6, 2004.

1. Johnson R, Frary C, Wang M. The nutritional conse-quences of flavored-milk consumption by school-agedchildren and adolescents in the United States. J AmDiet Assoc. 2002;102:853-856.

2. Oshiro C, Maskarinec G, Petitpain D, Hebshi S, No-votny R. Soy intervention in adolescent girls: Designand implementation. J Nutr Educ Behav. 2004;36:204-208.

3. Rivas M, Garay RP, Escanero JF, Cia P Jr, Cia P,Alda JO. Soy milk lowers blood pressure in men andwomen with mild to moderate essential hyperten-sion. J Nutr. 2002;132:1900-1902.

4. Bricarello LP, Kasinski N, Bertolami MC, Faludi A,Pinto LA, Relvas WGM, Izar MCO, Ihara SSM, TufikS, Fonseca FAH. Comparison between the effects ofsoy milk and nonfat cow milk on lipid profile and lipidperoxidation in patients with primary hypercholes-terolemia. Nutrition. 2004;20:200-204.

5. McArthur LH, Anguiano R, Gross KH. Are householdfactors putting immigrant Hispanic children at riskof becoming overweight: A community-based study ineastern North Carolina. J Community Health. 2004;

29:387-404.

April 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION 593