C

MFA

AO4cvcDfisifsqSetyfiIinuCbMmFbSv

ACtSmagPs

RmBs

A

©

RESEARCH

urrent Research

otivating 18- to 24-Year-Olds to Increase Theirruit and Vegetable Consumption

MY RICHARDS, MS, RD; KENDRA K. KATTELMANN, PhD, RD; CUIRONG REN, PhD

fPFRigb0i2CcJ

Cw2gtBlbhsipf

aasSgac

csrgfao

oqub

BSTRACTbjective This study assessed the effectiveness of a-month intervention using stage-based newsletters,omputer-based communication, and motivational inter-iewing to increase fruit and vegetable consumption byollege students aged 18 to 24 years.esign Participants were stratified by stage of change forruit and vegetable consumption and randomized to anntervention or control group. Participants completed thetaging algorithm for fruit and vegetable intake, whichncluded a one-item food frequency question, a 26-itemood frequency questionnaire (FFQ), an 18-item deci-ional balance questionnaire, and a five-item self-efficacyuestionnaire at baseline and completion of study.ubjects A convenience sample of 437 college studentsnrolled in a rural, land grant university was enrolled inhe study. Only nondietetics majors between ages 18 to 24ears were included in the study. A total of 314 studentsnished the study for a completion rate of 72%.

ntervention After baseline staging and randomization, thentervention group participants received four stage-basedewsletters, one motivational interview, and an individ-ally tailored e-mail follow-up over a 4-month period.ontrol group participants only received assessment ataseline and at completion.ain outcome measures Two fruit and vegetable instru-ents, a one-item food frequency question, and a 26-itemFQ measured daily consumption of fruits and vegeta-les at baseline and postintervention.tatistical analyses performed The SAS system for Windows,ersion 8 (1999, SAS Institute, Inc, Cary, NC), was used

. Richards is a child nutrition program specialist,hild and Adult Nutrition Services, Pierre, SD; at the

ime of the study, she was a master’s degree student atouth Dakota State University, Brookings. K. K. Kattel-ann is an associate professor, Nutrition, Food Science,

nd Hospitality Department, and director, Didactic Pro-ram, and C. Ren is an assistant professor of statistics,lant Science Department, South Dakota State Univer-ity, Brookings.

Address correspondence to: Kendra Kattelmann, PhD,D, Nutrition, Food Science, and Hospitality Depart-ent, South Dakota State University, Box 2275A,rookings, SD 57707. E-mail: Kendra.Kattelmann@

dstate.eduCopyright © 2006 by the American Dietetic

ssociation.0002-8223/06/10609-0015$32.00/0

udoi: 10.1016/j.jada.2006.06.005

2006 by the American Dietetic Association

or analysis, including the following tests: PROC GLM,ROC FREQ, and PROC NPAR1WAY, Kruskal-Wallis,isher, Wilcoxon rank sum, and �2.esults Fruit and vegetable consumption increased signif-cantly more for the intervention group than the controlroup. Consumption increased in the intervention groupy one serving a day for both instruments compared with.4 servings a day in the control group for a one-itemnstrument and no change in the control group for a6-item FFQ.onclusions. This intervention is an effective way to in-rease fruit and vegetable consumption by young adults.Am Diet Assoc. 2006;106:1405-1411.

ollege students aged 18 to 24 years are at an age oftransitioning from parental supervision to indepen-dent living and are developing food patterns that

ill affect their future. In 2000, more than 35% of 18- to4-year-olds were enrolled in college (1). Generally inood health, young adults often are ambivalent aboutheir future health and the role that nutrition plays (2).ecause of their lack of medical problems, there has been

ittle focus on 18- to 24-year-olds. It can take decadesefore diet-related disease appears. A strong associationas been established between fruit and vegetable con-umption and a decreased risk of chronic diseases (3-9). Its estimated that 20% or more cases of cancer could berevented by consuming a diet that contains a variety ofruits and vegetables in substantial amounts (10).

Many young adults are not consuming enough fruitsnd vegetables to prevent diet-related disease. Nation-lly, the percentage of people consuming five or moreervings of fruits and vegetables a day is only 23.1%; inouth Dakota it is only 19.9% (11,12). Finding a way toet young people to eat more fruits and vegetables couldssist them in preventing and delaying overweight andhronic diseases such as cancer and hypertension.Making the choice to increase their fruit and vegetable

onsumption is a simple and inexpensive preventiontrategy that 18- to 24-year-olds can use to reduce theirisk of chronic disease. Interventions aimed at this ageroup have the potential added benefit of influencinguture generations because 18- to 24-year-olds are at ange when many start families and pass nutrition habitsn to their children.The intent of this study was to motivate 18- to 24-year-

lds to consume more fruits and vegetables. The researchuestion that was addressed in this study was: “Will these of stage-based newsletters with the addition of stage-ased motivational interviewing, computer-based follow-

p, and a nutrition Web site increase fruit and vegetable

Journal of the AMERICAN DIETETIC ASSOCIATION 1405

cs

MTvlt2vdvidyapvogofsqD

tisiltmwtsPttr(r6a

hyvtlctbcocmoa

cr3

mtf

prsqtt

tnvetpcplbt

bttfnafrpfcatf

tcpp

cpivhtfirl

cvdpsco(

1

onsumption by 18- to 24-year-old Midwestern collegetudents?”

ETHODShis study assessed the effectiveness of a 4-month inter-ention (January through April) using stage-based news-etters, computer-based communication, and motiva-ional interviewing on a convenience sample of 437 18- to4-year-old college students from a rural land grant uni-ersity with an approximate enrollment of 10,900 stu-ents. The study was approved by the Institutional Re-iew Board before recruitment of participants. To bencluded in the study, participants had to be college stu-ents; be nondietetics majors; be between ages 18 and 24ears; sign a consent form; and have a current e-mailddress, mail address, and telephone number. Partici-ants were stratified by stage of change for fruit andegetable consumption and randomized to interventionr control group. Participants completed the staging al-orithm for fruit and vegetable intake, which included ane-item food frequency question (Measure 1), a 26-itemood frequency questionnaire (FFQ), an 18-item deci-ional balance questionnaire, and a five-item self-efficacyuestionnaire at baseline and at completion of the study.emographic information was collected at baseline.Stage of change was measured using an algorithm (13)

hat based stage assignment upon whether or not thendividual perceived that he or she consumed five or moreervings of fruits and vegetables per day. Participantsndicated how many servings a day they consumed and ifess than five servings a day, indicated how soon theyhought they would increase their consumption to five orore servings of fruits and vegetables. If individualsere consuming five or more servings of fruits and vege-

ables, the algorithm determined if they had been doingo for fewer than 6 months or more than 6 months.articipants were considered as precontemplation (no in-ention of changing in the next 6 months), precontempla-ion (thinking of changing in the next 6 months), prepa-ation (planning to change in the next 30 days), actionhave already changed behavior to meet a specific crite-ion), and maintenance (behavior has been sustained formonths) (14-18) depending on their response to the fruitnd vegetable algorithm.The 26-item FFQ (19) asked how often the individual

ad eaten specific fruits and vegetables during the pastear. The questionnaire included 12 fruit items and 14egetable items with “how often” and “how much” quan-ifiers as well as portion choices of small, medium, andarge. Measure 1 was part of a staging algorithm andonsisted of a one-item question that asked respondentshe question: “How many servings of fruits and vegeta-les do you usually eat each day? (a serving is ½ cup ofooked or raw vegetables, 1 cup of salad, a piece of fruit,r ¾ cup [6 ounces] of 100% fruit juice).” The answerhoices were zero, one, two, three, four, five, and six orore. The number six was assigned to the answer choice

f six or more per day for purposes of calculating aver-ges.Decisional balance (19) was measured at baseline and

ompletion of study using a scale with 18 items that wereated from one to five, where 1�strongly disagree,

�neutral, and 5�strongly agree. Decisional balance was i

406 September 2006 Volume 106 Number 9

easured separately for fruit and vegetable consumptiono determine the balance of pros vs cons for consumingruits and for consuming vegetables.

Self-efficacy (19) was measured at baseline and at com-letion of the study with five items using a Likert scaleanging from one to nine. Self-efficacy was measuredeparately for fruits and vegetables. The self-efficacyuestions measured how confident the participants werehat they could consume two to four servings of fruits andhree to five servings of vegetables a day.

The intervention group received a personalized letterailored to stage of change, a series of four stage-basedewsletters specific to their stage at baseline, one moti-ational-interviewing session and a minimum of two-mail contacts during the 4-month intervention. One ofhe e-mail contacts informed the intervention partici-ants of the address of a nutrition Web site designed forollege students. The intervention participants who com-leted a motivational interview also received tailored fol-ow-up by e-mail. The control group only completed theaseline and completion surveys with no additional con-act from study personnel during the study.

Stage-based newsletters for fruits and vegetables com-ined were developed by Geoffrey Greene, PhD, RD, athe University of Rhode Island and permission was ob-ained for use in this intervention. There was a series ofour newsletters for each of the five stages of change. Theewsletters were tailored to each stage and used stageppropriate processes of change. They were printed inull color with visually interesting type and photographs,ecipes, and a question-and-answer section. Precontem-lators and contemplators were given reasons to eat moreruits and vegetables and preparation newsletters in-luded goal setting and tips for consuming more fruitsnd vegetables. Action and maintenance newsletters con-ained tips for maintaining consumption and trying newruits and vegetables.

The Web site was developed by the principal investiga-or and contained recipes, serving size guidelines, a costomparison of a meal with fruits and vegetables com-ared with the cost of a fast-food meal, nutrition facts,reparation tips, and links to other nutrition Web sites.Motivational interviewing was completed by the prin-

ipal investigator. The same questions were used for allarticipants based on previous validated motivationalnterviewing by Resnicow and colleagues (20). The inter-iew was designed to help the participant recognize his orer own specific barriers to fruit and vegetable consump-ion and develop possible solutions to the barriers. E-mailollow-up was tailored to the participant’s motivationalnterview. Examples of e-mail message content includedecipes, fun nutrition facts, fruit and vegetable tips, andinks to nutrition Web sites for more information.

To determine if there was a difference between thoseonsuming five or more servings per day of fruits andegetables and those consuming less than 5 servings peray, the data from precontemplation, contemplation, andreparation were combined and designated the Preactiontage and the data from action and maintenance wereombined and designated the Action stage. The groupingsf Preaction and Action have been reported previously21,22). This type of grouping determines where a partic-

pant is in relation to the desired action. Either the par-

ta

pCiPirrdnasa

RTabc

cdwnd4(paciiWTfgto

ftmtts1wptnviP0c

icipant is doing the action (Action) or not doing thection (Preaction).Statistical analysis was completed by using several

rocedures in SAS (version 8.0, 1999, SAS Institute, Inc,ary, NC). Procedures used for the parametric statistics

ncluded: PROC GLM, PROC FREQ, and PROC REG.rocedures for nonparametric statistics used for categor-

cal and non-normal data included: �2, Fisher, Wilcoxonank sum, and Kruskal-Wallis. The null hypothesis wasejected at the 0.05 level of significance. Normality for theata was tested. Transformations were conducted foron-normal data including logarithmic transformationnd the transformed data satisfied the normality as-umptions. Test statistics included �2, Student t tests,nd F tests.

ESULTShere was not a significant difference between the controlnd intervention groups for demographic measures ataseline (Table 1). The average age of participants who

Table 1. Demographics of 18- to 24-year-old college students(n�314) who completed a study that assessed the effectiveness ofa 4-month intervention using stage-based newsletters, computer-based communication, and motivational interviewing to increasefruit and vegetable consumptiona

Demographic

Control(n�157)(n)

Intervention(n�157)(n)

Total(n)

% oftotal

SexMen 35 43 78 24.8Women 122 114 236 75.2

Student statusFull-time students 154 152 306 97.8Part-time students 2 5 7 2.2

RaceWhite, non-Hispanic 153 151 304 96.8African American,

non-Hispanic 0 0 0 0Hispanic/Latino, Latina 0 1 1 0.3American Indian 0 1 1 0.3Asian/Pacific Islander 4 4 8 2.5Other 0 0 0 0

Rural vs nonruralSpent �80% life on

farm/ranch 38 45 83 27.5Spent �79% life on

farm/ranch 116 103 219 72.5

Living situationLives in dorms 62 61 123 39.2Does not live in dorms 95 96 191 60.8

aThere were no differences in demographics between control and intervention partic-ipants (�2, P�0.11).

ompleted the project was 20.4�1.5 years. Although a s

S

onvenience sample was recruited, there was a uniformistribution among colleges within the university fromhich the participants were recruited. There was no sig-ificant difference in the numbers of students from theifferent colleges at the university (�2, P�0.93). Of the37 baseline participants, 310 (71%) were women and 12729%) were men. Three hundred fourteen (72%) com-leted the study. The dropout rate was 24% for womennd 39% for men, resulting in 236 women and 78 menompleting the study (Table 1). There were no differencesn demographics between completers and noncompletersn the study except for sex (Kruskal-Wallis P�0.001).

omen completed the study at a higher rate than men.here were no differences in baseline or postintervention

ruit and vegetable consumption based on the demo-raphics measured, including sex, rural vs nonrural sta-us, whether or not students lived in dorms, and whetherr not they had a campus meal plan.The distribution of the number of servings per day of

ruits and vegetables from the 26-item FFQ was posi-ively skewed. Data were transformed using log transfor-ations for statistical analyses. The values presented are

he actual values, and the P values reported are based onhe transformed data (Table 2). Fruit and vegetable con-umption from baseline to postintervention increased by.0�0.3 servings a day (paired t test P�0.001) comparedith no change in consumption (0.1�0.3 servings a day,aired t test P�0.48) in the control group as estimated byhe 26-item FFQ. Data from Measure 1 approximated aormal distribution and were not transformed. Fruit andegetable consumption from baseline to postinterventionncreased by 1.0�0.1 servings a day (paired t test�0.001) in the intervention group compared with.4�0.1 servings a day (paired t test P�0.001) in theontrol group as estimated by Measure 1. For both mea-

Table 2. Comparison of fruit and vegetable consumption by 18- to24-year-old college students (n�157 control, n�157 intervention)at baseline and postintervention with stage-based newsletters,computer-based communication, and motivational interviewing

Measure

Baselineservings/day(mean�SEa)

Postservings/day(mean�SE)

Changeservings/day(mean�SE)

1-Item FFQb

Control 2.1�0.1 2.5�0.1 0.4�0.1d

Intervention 2.2�0.1 3.2�0.1 1.0�0.1d

26-Item FFQc

Control 5.2�0.3 5.2�0.3 0.0d

Intervention 5.4�0.4 6.3�0.4 1.0�0.3d

aSE�standard error.bFFQ�food frequency questionnaire. The one-item food frequency question wasadministered in the staging algorithm.cThe 26-item FFQ assessed fruit and vegetable intake. Data were log transformed forstatistical analyses. P values are for transformed data. Data reported are actual data.dSignificant difference between intervention and control group for change in servings/day (one-item question: Wilcoxon two sample test, P�0.001; 26-item FFQ: analysis ofvariance, P�0.04).

ures the intervention group had significantly greater

eptember 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION 1407

ic

(va2tKwM(

ptptcaftpc

c

tuwrite(dtb

vwhtitims

saivvp

CTnfscdtbsmtfc

ttegcddsesnbbmr2

m

1

ncreases in fruit and vegetable consumption than theontrol group (Table 2).

When the data were grouped into Preaction and Actionbased on baseline stage of change), the Preaction inter-ention group had a significantly greater increase in fruitnd vegetable consumption for both Measure 1 and the6-item FFQ for change in fruit and vegetable consump-ion (Measure 1, Kruskal-Wallis P�0.001; 26-item FFQ,ruskal-Wallis P�0.03) than the control group. Thereere no differences between control and intervention foreasure 1 (Kruskal-Wallis P�0.94) and the 26-item FFQ

Kruskal-Wallis P�0.71) for the Action group (Table 3).The number of participants by stage at baseline and

ostintervention is shown in the Figure. Both interven-ion and control groups decreased in number of partici-ants in precontemplation from baseline to postinterven-ion. The movement among the stages was analyzed as aategorical variable. The participants were randomizednd stratified by stage to their group at baseline; there-ore there was no difference between stages of change forhe two groups at baseline. At completion the number ofarticipants by stage was significantly different betweenontrol and intervention (�2 17.8, P�0.01).Decisional balance scores (pros and cons) and self-effi-

Table 3. Fruit and vegetable intake of 18- to 24-year-old collegestudents by Preaction and Action grouping

Group

Baselineservings/day(mean�SEa)

Completionservings/day(mean�SE)

Preactionb

1-Item FFQc

Control 1.9�0.1 2.4�0.1d

Intervention 2.0�0.1 3.1�0.1

26-Item FFQe

Control 5.0�0.3 5.1�0.3d

Intervention 4.9�0.3 6.1�0.4

Actionb

1-Item FFQControl 5.1�0.1 4.4�0.3Intervention 5.3�0.2 4.6�0.3

26-Item FFQe

Control 8.3�1.6 7.0�1.0Intervention 12.2�3.9 9.9�2.7

aSE�standard error.bTo determine if there was difference between those consuming five or more servingsof fruits and vegetables/day and those consuming less than five servings/day data fromprecontemplation, contemplation, and preparation stage groups were combined anddesignated Preaction and data from action and maintenance stage groups werecombined and designated Action.cFFQ�food frequency questionnaire.dIntervention Preaction group increased fruit and vegetable consumption more than thecontrol Preaction group (26-item FFQ: Kruskal-Wallis, P�0.03; one-item FFQ: Kruskal-Wallis, P�0.001).eData were log transformed for statistical analyses. P values are for transformed data. Datareported are actual data.

acy scores did not differ between control and interven- m

408 September 2006 Volume 106 Number 9

ion at baseline. At completion, covariance adjusted meansing baseline value for each construct as the covariateas used for analysis and presentation to more accu-

ately reflect change during the intervention period. Thentervention self-efficacy scores for both fruits and vege-ables were significantly greater than control group self-fficacy scores at completion. Decisional balance scorespros and cons) for both fruits and vegetables neitheriffered significantly from baseline to completion nor be-ween intervention and control groups at completion (Ta-le 4).Participants were asked on the final survey if they had

isited the Web site. Of the 153 intervention participantsho answered that question, 80 (52.3%) stated that theyad visited the Web site and 70 (45.8%) indicated thathey had not visited the Web site. Three participantsndicated that they did not remember if they had visitedhe Web site. Those students who visited the Web site didncrease their fruit and vegetable consumption slightly

ore than those who did not visit the Web site by Mea-ure 1 (P�0.001) but not by the 26-item FFQ.Of the 157 intervention participants who completed the

tudy, 150 (95.5%) received motivational interviewingnd seven (4.5%) did not receive motivational interview-ng. There was not a significant difference in fruit andegetable consumption for those who completed a moti-ational interview, set goals, or replied to e-mail com-ared with those who did not complete these activities.

ONCLUSIONShis study demonstrated that the use of four stage-basedewsletters, one motivational interview, and an e-mailollow-up intervention increased fruit and vegetable con-umption of 18- to 24-year old college students. Whenomparing those who said that they ate five servings aay (action and maintenance stages) to those who saidhey did not eat five servings a day of fruits and vegeta-les (precontemplation, contemplation, and preparationtages) the intervention group had a significantly greaterovement to action than the control group. This suggests

hat this intervention was successful at moving peoplerom Preaction stages of change to Action stages ofhange.Both measures of fruit and vegetable intake used in

his study showed a greater increase in consumption inhe intervention group than in the control group. How-ver, the intakes as measured by the 26-item FFQ werereater than five servings a day both at baseline and atompletion (Table 2). We were using the 26-item FFQ toocument changes in intake due to the intervention so weo not view this as a limitation in interpretation of re-ults. Furthermore, FFQs have estimated fruit and veg-table intake near five servings a day or more than fiveervings a day in other studies (23-28). In addition, theumber of fruit and vegetable questions on an FFQ haseen found to influence the number of fruits and vegeta-les that are estimated to be consumed per day (29). Thisay be one reason that the one-item measure in this

esearch study estimated intake much lower than the6-item measure.The lack of significant differences in decisional balanceay be due to measurement incongruence between the

easurement tools for decisional balance and the interven-

tw(

bb

tlts

tcadrscfafsac

Ftic ch ot

ion. The assessment for decisional balance and self-efficacyas based on fruits (target of two servings) and vegetables

Table 4. Decisional balance (pros and cons) and self-efficacyscores at completion for 18- to 24-year-old college students whocompleted a study that assessed the effectiveness of a 4-monthintervention using stage-based newsletters, computer-based com-munication, and motivational interviewing to increase fruit andvegetable consumptiona

Decisionalbalance

Control (n�157) Intervention (n�157)

Fruits Vegetables Fruits Vegetables

4™™™™™™™™™™ mean�standard error ™™™™™™™™™3Pros 32.9�0.3 32.6�0.4 33.6�0.3 33.1�0.3Cons 24.7�0.4 25.1�0.4 24.6�0.4 25.2�0.4Self-efficacy 33.0�0.5 30.8�0.6 34.4�0.5b 32.0�0.6b

aBaseline value for each construct were used as the covariate.bMean self-efficacy scores are significantly different between control and interventionfor respective fruits and vegetables (P�0.05).

igure. Number of participants by stage at baseline and completion fohe effectiveness of a 4-month intervention using stage-based newsncrease fruit and vegetable consumption. aParticipants were stratifiedompletion, control and intervention are significantly different from ea

target of three servings) separately, whereas the stage- s

S

ased intervention was based on fruits and vegetables com-ined with a target of five servings per day.A limitation of this study is that there was no follow up

o evaluate the effectiveness of the intervention at regu-ar intervals over the years. An additional limitation washe use of a convenience sample rather than a randomample.This age group may have a tendency to fluctuate in

heir fruit and vegetable consumption depending onhanging circumstances, such as moving off campusnd/or cooking for themselves. Future research could beirected at discovering how young adults’ changing envi-onment as they transition from parents, to peers, topouses and children affects their fruit and vegetableonsumption. Because men were more difficult to recruitor this study and because men dropped out of the studyt a higher rate than women, future research could beocused on sex differences in fruit and vegetable con-umption and different communication strategies for mennd women to increase and maintain fruit and vegetableonsumption.Results from this intervention suggest that use of

to 24-year-old college students who completed a study that assesseds, computer-based communication, and motivational interviewing totage and randomized to control or intervention group at baseline. bAther, �2�17.8, P�0.01.

r 18-letterby s

tage-based newsletters in conjunction with a motiva-

eptember 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION 1409

tvtwpirop

TmfUs

tS

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

1

ional interview and e-mail contact increases fruit andegetable consumption in young adults. Once newslettersailored to stage of change were developed and a Web siteas established, a health professional on a college cam-us could implement and run this intervention for a sim-lar number of students per semester. It is a practical andelatively inexpensive intervention that has the potentialf not only improving the long-term health of young peo-le, but also their future children as well.

his project was partially funded by Agriculture Experi-ent Station, South Dakota State University. Additional

unding was obtained from an Alumni Grant from the Phipsilon Omicron Foundation and the University Re-

earch Support, South Dakota State University.This article was published with approval of the direc-

or, Agricultural Experiment Station, South Dakotatate University, as publication #3514 of a journal series.

eferences1. US Census Bureau. Table A-5. The population 14 to

24 years old by high school graduate status, collegeenrollment, attainment, sex, race, and Hispanic ori-gin: October 1967 to 2000. 2001. Available at: http://www.census.gov/ftp/pub/population. Accessed Au-gust 2, 2002.

2. Betts NM, Amos RJ, Georgiou C, Hoerr SL, Ivaturi R,Keim KS, Tinsley A, Voichick J. What young adultssay about factors affecting their food intake. EcolFood Nutr. 1995;34:59-64.

3. Lampe JW. Health effects of vegetables and fruit: As-sessing mechanisms of action in human experimentalstudies. Am J Clin Nutr 1999;70(suppl 3):475S-490S.

4. Hu FB, Rimm EB, Stampfer MJ, Ascherio A,Speigelman D, Willett WC. Prospective study of ma-jor dietary patterns and risk of coronary heart dis-ease in men. Am J Clin Nutr. 2000;72:912-921.

5. Smith-Warner SA, Elmer PJ, Tharp TM, Fosdick L,Randall B, Gross M, Wood J, Potter JD. Increasingvegetable and fruit intake: Randomized interventionand monitoring in an at-risk population. Cancer Epi-demiol Biomarkers Prev. 2000;9:307-317.

6. Liu S, Manson JE, Lee I-M, Cole SR, Hennekens CH,Willett WC, Buring JE. Fruit and vegetable intakeand risk of cardiovascular disease: the Women’sHealth Study. Am J Clin Nutr. 2000;72:922-928.

7. Joshipura KJ, Ascherio A, Manson JE, Stampfer MJ,Rimm EB, Speizer FE, Hennekens CH, Spiegelman D,Willett WC. Fruit and vegetable intake in relation torisk of ischemic stroke. JAMA. 1999;282:1233-1239.

8. Van Duyn MAS, Pivonka E. Overview of the healthbenefits of fruit and vegetable consumption for thedietetics professional: Selected literature. J Am DietAssoc. 2000;100:1511-1521.

9. Hwang MY. Why you should eat more fruits andvegetables. JAMA. 1999;282:1304.

0. World Cancer Research Fund and American Institutefor Cancer Research. Food, Nutrition and the Preven-tion of Cancer: A Global Perspective. Washington, DC:American Institute for Cancer Research; 1997.

1. Centers for Disease Control and Prevention. Behavioral

Risk Factor Surveillance System Survey Data. Prev-

410 September 2006 Volume 106 Number 9

alance Data: Nationwide and South Dakota. “What isyour average frequency of fruit and vegetable consump-tion per day?” Available at: http://apps.nccd.cdc.gov/brfss/display.asp?cat�NU&yr�2000&qkey�4340&state�US. Accessed July 31, 2002.

2. South Dakota Department of Health. BehavioralRisk Factor Surveillance System: 1999 South DakotaSummary. Available at: http://www.state.sd.us/doh/Pubs2/1999BRFSS.doc. Accessed July 31, 2002.

3. Ma J, Betts NM, Horacek TM. Measuring stage ofchange for assessing readiness to increase fruit andvegetable intake among 18- to 24-year-olds. Am JHealth Promot. 2001;16:88-97.

4. Finckenor M, Byrd-Bredbenner C. Nutrition inter-vention group program based on preaction-stage-ori-ented change processes of the Transtheoretical Modelpromotes long-term reduction in dietary fat intake.J Am Diet Assoc. 2000;100:335-342.

5. Greene GW, Rossi SR, Reed GR, Willey C, ProchaskaJO. Stages of change for reducing dietary fat to 30%of energy or less. J Am Diet Assoc. 1994;94:1105-1110.

6. Campbell MK, DeVellis BM, Strecher VJ, Ammer-man AS, DeVellis RF, Sandler RS. Improving dietarybehavior: The effectiveness of tailored messages inprimary care settings. Am J Public Health. 1994;84:783-787.

7. Prochaska JO, Velicer WF. The TranstheoreticalModel of Health Behavior Change. Am J Health Pro-mot. 1997;12:38-48.

8. Prochaska JO, Velicer WF, Rossi JS, Goldstein MG,Marcus BH, Rakowski W, Fiore C, Harlow LL, Red-ding CA, Rosenbloom D, Rossi SR. Stages of changeand decisional balance for 12 problem behaviors.Health Psychol. 1994;13:39-46.

9. Ma J, Betts NM, Horacek T, Georgiou C, White A,Nitzke S. The importance of decisional balance andself-efficacy in relation to stages of change for fruitand vegetable intakes by young adults. Am J HealthPromot. 2002;16:157-166.

0. Resnicow K, Jackson A, Wang T, Anindya KD, Mc-Carty F, Dudley WN, Baranowski T. A motivationalinterviewing intervention to increase fruit and vege-table intake through Black churches: Results of theEat for Life Trial. Am J Public Health. 2001;91:1686-1693.

1. Walton J, Hoerr S, Heine L, Frost S, Roisen D,Berkimer M. Physical activity and stages of change infifth and sixth graders. J Sch Health. 1999;69:285-289.

2. Vallis M, Ruggiero L. Predictors of successful behaviorchange for healthy eating: Diabetes Stages of Change(DiSC) Study [abstract]. Diabetes. 2003;52:A420.

3. Warneke CL, Davis M, De Moor C, Baronowski T. A7-item versus 31-item food frequency questionnaire formeasuring fruit, juice, and vegetable intake among apredominantly African-American population. J Am DietAssoc. 2001;101:774-779.

4. Thompson FE, Subar AF, Smith AF, Midthune D,Radimer KL, Kahle LL, Kipnis V. Fruit and vegeta-ble assessment: Performance of 2 new short instru-ments and a food frequency questionnaire. J Am Diet

Assoc. 2002;102:1764-1772.

2

2

2

2

2

5. Field AE, Colditz GA, Fox MK, Byers T, Serdula M,Bosch RJ, Peterson KE. Comparison of 4 question-naires for assessment of fruit and vegetable intake.Am J Public Health. 1998;88:1216-1218.

6. Thompson FE, Kipnis V, Subar AF, Krebs-Smith SM,Kahle LL, Midthune D, Potischman N, Schatzkin A.Evaluation of 2 brief instruments and a food-fre-quency questionnaire to estimate daily number ofservings of fruit and vegetables. Am J Clin Nutr.2000;71:1503-1510.

7. Stevens VJ, Glasgow RE, Toobert DJ, Karanja N,

Smith SK. Randomized trial of a brief dietary inter-

S

vention to decrease consumption of fat and increaseconsumption of fruits and vegetables. Am J HealthPromot. 2002;16:129-134.

8. Harris J, Murray L. The relationship between socialcognitive model variables and fruit and vegetableconsumption among college students. Am J HealthStudies. 1997;13:133-140.

9. Krebs-Smith SM, Heimendinger J, Subar AF, Patter-son BH, Pivonka E. Using food frequency question-naires to estimate fruit and vegetable intake: Associ-ation between the number of questions and total

intakes. J Nutr Educ. 1995;27:80-85.

eptember 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION 1411