Evaluating changeability to improve fruit and vegetable intake among school aged children
© Nanney et al; licensee BioMed Central Ltd. 2005
Received: 07 September 2005
Accepted: 21 November 2005
Published: 21 November 2005
The purposes of this paper are two fold. First, to describe an approach used to identify fruits and vegetables to target for a child focused dietary change intervention. Second, to evaluate the concept of fruit and vegetable changeability and feasibility of applying it in a community setting.
Steps for identifying changeable fruits and vegetables include (1) identifying a dietary database (2) defining geographic and (3) personal demographics that characterize the food environment and (4) determining which fruits and vegetables are likely to improve during an intervention. The validity of these methods are evaluated for credibility using data collected from quasi-experimental, controlled design among 7–9 year old children (n = 304) participating in a tutoring or mentoring program in St. Louis, MO. Using a 28-item food frequency questionnaire, parents were asked to recall for their child how often foods were eaten the past 7 days. This questionnaire was repeated eight months later (response rate 84%). T-test analyses are used to determine mean serving differences from baseline to post test.
The mean serving differences from baseline to post test were significant for moderately eaten fruits (p < .001), however, not for vegetables (p = .312). Among the intervention group, significantly more children ate grapes (p < .001), peaches (p = .022), cantaloupe (p < .001), and spinach (p = .044) at post testing – all identified as changeable with information tailored to participants.
Data driven, food focused interventions directed at a priority population are feasible and practical. An empirical evaluation of the assumptions associated with these methods supports this novel approach. However, results may indicate that these methods may be more relevant to fruits than vegetables. This process can be applied to diverse populations for many dietary outcomes. Intervention strategies that target only those changeable fruits and vegetables are innovative and warrant further study.
Children in the United States follow eating patterns that do not meet national recommendations for fruits and vegetables. [1, 2] To date, dietary interventions have been only modestly successful in increasing fruit and vegetable (FV) intake among children  with effect sizes ranging from .2 servings (Gimme 5)  to .99 servings (High 5 Project).  One means of improving the impact of dietary interventions is to assure the intervention is relevant to a child by targeting foods that are available and accessible in their environment. 
Program planning models suggest that directing program resources toward those factors that are most changeable helps to ensure program efficiency and effectiveness. [7, 8] For dietary interventions, the concept of changeability can be defined as identifying those FV that have the greatest likelihood for increased consumption and targeting them for an intervention. Changeability is moderated by variations in food consumption patterns across racial and ethnic, [9–12] gender, [9, 10, 12–15] age, [11, 13, 15] marital status,  and regional [11, 13, 16] differences. Changeability is also influenced by the discrepancies in the availability of food in the community and home environments, particularly in priority populations. [17–19]
From an individual perspective, the food experiences to which young children are exposed are critical to the early development of food acceptance patterns and choices. This exposure to FV influences familiarity, preferences, and intake. For purposes of an intervention, FV that are preferred and eaten by a child are less changeable because consumption goals are met. Instead, moderate intake of FV suggests foods that are familiar, accessible, and changeable. Systematic approaches derived from program planning constructs, are needed to determine program relevance and changeability in order to successfully impact health behaviors. 
The purpose of this paper is to describe a systematic approach used for identifying changeable FV that were targeted for intervention in Partners of all Ages Reading About Diet and Exercise (PARADE), a school based mentoring program designed to improve FV intake of children ages 7 to 9 years. Children enrolled in the mentoring program received one to one tutoring on a weekly basis during regular school hours. PARADE was incorporated within the routine curriculum of the mentoring program. PARADE was based on social cognitive theory and included (1) eight lesson plans with computer-generated storybooks tailored to the dietary patterns of each 7–9 year old child and (2) eight mailed parent newsletters introducing each book and offering tips on how to role model healthy eating at home. PARADE was evaluated using a quasi-experimental study design. Participants in the evaluation include children and their parents. Approval for this study was obtained from the Saint Louis University Institutional Review Board.
Steps for identifying changeableFV
Limited time and resources increased the importance of identifying appropriate foods for change in this priority population. The following steps were used to identify and tailor information on FV as part of PARADE lesson plans and storybooks.
Step 1: Identify a comprehensive nutrition database
Dietary measurement is complex. Accuracy can be maximized with methods that help individuals correctly recall the foods and amounts eaten on any given day that represents usual intake. Furthermore, a comprehensive list of foods and nutrient values is essential to calculating the dietary outcomes of interest.  The Continuing Survey of Food Intakes by Individuals (CSFII) database represents data from the priority population and is methodologically and psychometrically sound. Also popularly known as the "What We Eat in America Survey," the CSFII is a national food consumption survey conducted by the Agricultural Research Service of the United States Department of Agriculture.  The 10th edition of the CSFII (1994–1996) provides nationally representative data by over sampling for low-income individuals and young children. Individuals are asked to provide food intakes on two nonconsecutive days using a multiple pass 24-hour dietary recall administered in the home by trained interviewers. Sample sizes include 12,700 adults of all ages and 11,800 children birth-19 years. Response rates include 1-day 80% and 2-day 76%. Food consumption data are available on CD-ROM from the USDA. More information, including ordering the CD's, can be found at http://www.ars.usda.gov/Services/docs.htm?docid=7787.
Step 2: Define the sociodemographic characteristics of the priority population
Variations in dietary patterns are influenced by individual characteristics such as race, [9–12] gender, [9, 10, 12–15] and age. [11, 13, 15] Thus, it is critical to assess these factors with regard to determining changeability of fruits and vegetables. For the CSFII database, food consumption differences can be viewed by five race categories; white, black, Pacific Asian, Native American, and Other. Furthermore, the data can be segmented by gender and age. The CSFII database reports age in months for children less than one year of age, and in years for those over 1 year of age. For Project PARADE, the CSFII data was analyzed for children 7–9 years of age (9% of CSFII sample), male (49%) and female (51%), and African American (18%). 892 children matched these criteria.
Step 3: Define the geographic characteristics of the priority population
Dietary intake varies by geographic region and urban versus rural influences. The CSFII divides the United States into five geographic regions that are further defined by urbanization type. Urbanization types include city, outside the city, and rural areas. PARADE participants lived in ten urban and suburban counties in a large Midwest city. Therefore, the CSFII data was analyzed for those living in a metropolitan city and outside the city in the Midwest. Results identified nearly three-quarters of the sample living in a metropolitan area (central city, 21% and suburban, 53%). These geographic criteria resulted in a final sample of dietary intake data for 164 children, mean age 7.93 (SD = .82). Steps 1–3 identified 2423 food entries by the CSFII. This included 670 unique foods including 49 fruits and 79 vegetables consumed by the sample.
Step 4: Identify changeable FV for the priority population
Steps to identify relevant foods to target in a dietary intervention
Steps to Determine Foods for a Dietary Intervention
Key Characteristics that Effect Dietary Intake
Define Sociodemographic Variables
• Married, Divorced, Separated, Single, Widowed
• Caucasian, African American, Pacific Asian, Native American, and Other
• Male, Female
• 0–12 Months, Years
Define Geographic Variables
• North, South, East, West, Midwest
• City, Suburban, Rural
Identify Changeable Foods to Target
• Rank % between 25 and 75
Using national frequency data to estimate local FV consumption
Fruits and vegetables identified as changeable and targeted for intervention
CSFII Weighted % Consumed1 (n = 502)
CSFII Rank %
PARADE Baseline % Consumed2 (n = 304)
PARADE Baseline Rank %
CF Fruit Cocktail
CSFII Weighted % Consumed1
CSFII Rank %
PARADE Baseline % Consumed2
PARADE Baseline Rank %
CV Green beans
CV Mixed Vegetables
CV Cabbage slaw
CV Green Peas
Fruits and vegetables targeted for Project PARADE improved at post intervention
Average Pre % consumed
Average Post % consumed
PF Fruit Cocktail
Average Pre % consumed
Average Post % consumed
PV Green beans
PV Mixed Vegetables
PV Cabbage slaw
PV Green Peas
Among the intervention group, significantly more children ate grapes (p < .001), peaches (p = .022), cantaloupe (p < .001), and spinach (p = .044) at post testing – all identified as changeable with information tailored to PARADE participants. As hypothesized, consumption of fruits and vegetables frequently and rarely eaten remained unchanged or were eaten more in two instances (potatoes, corn). Worth noting is that the majority of the vegetables remained unchanged at the end of the intervention. These results may indicate that these methods may be more relevant to fruits than vegetables.
Moderately eaten fruits improved at post intervention
PARADE Intervention (N = 304)
Mean Servings Baseline
Mean Servings Post
Frequently Eaten Fruits
Moderately Eaten Fruits
Rarely Eaten Fruits
Frequently Eaten Vegetables
Moderately Eaten Vegetables
Rarely Eaten Vegetables
Recent findings suggest that innovative research is necessary to broaden the traditional approach beyond increasing FV awareness and education.  Overall, this study reinforces these methods as a way to systematically identify FV to target for an intervention. Identifying and concentrating on the most changeable behavioral targets for interventions direct resources to where they will be most beneficial. Furthermore, greater specificity in program development simplifies the evaluation process (i.e., brief food frequency questionnaire). [23, 24]
Assessments of changeability across fruit and vegetable patterns can be made by defining person and place variables of the priority population, examining the frequency with which fruits and vegetables have been consumed, and identifying those that can be reasonably expected to change for targets of an intervention. Moreover, this approach takes into account important environmental influences upon dietary patterns. Haire-Joshu and Nanney describe individual food preference, cultural and familial influences, and home, school and community environments as having significant influences upon the food environment of children.  This process addresses eating behavior as a function of the varied food environments for a specific population, albeit, in a broader community based context.
We developed, applied, and tested a systematic, data based approach to assess changeability and specify fruits and vegetables for an intervention among underserved school aged children. This approach will allow for further clarity of intervention effects by targeting only those changeable fruits and vegetables for intervention. Furthermore, this process can be applied to diverse populations for a variety of dietary outcomes. Additionally, larger mass media interventions like the 5 A Day campaign may benefit from this approach. More research is needed to evaluate the effectiveness and generalizability of community-based efforts that promote changeable foods for an intervention.
This work is based upon funding from the American Cancer Society (TURPG-00-286-01 PBP) and National Institutes of Nursing Research (USPHS-5-R01-NR05079). The authors would like to recognize Program Manager of the PARADE Project Brandye L. Mazdra and our community partners; The OASIS Institute, St. Louis area Big Brothers Big Sisters, and Girls Inc.
- Krebs-Smith SM, Cook A, Subar AF, Cleveland L, Friday J, Kahle LL: Fruit and Vegetable Intakes of Children and Adolescents in the United States. Arch Pediatr Adolesc Med. 1996, 150: 81-86.View ArticlePubMedGoogle Scholar
- Munoz KA, Krebs-Smith SM, Ballard-Barbash RB, Cleveland LE: Food Intakes of US Children and Adolescents Compared With Recommendations. Pediatrics. 1997, 100 (9): 323-329. 10.1542/peds.100.3.323.View ArticlePubMedGoogle Scholar
- Ammerman A, Lindquist C, Lohr KN, Hersey J: The Efficacy of Behavioral Interventions to Modify Dietary Fat and Fruit and Vegetable Intake: A Review of the Evidence. Preventive Medicine. 2002, 35: 25-41. 10.1006/pmed.2002.1028.View ArticlePubMedGoogle Scholar
- Baranowski T, Davis M, Resnicow K, Baranowski J, Doyle C, Lin LS, Smith M, Wang DT: Gimme 5 fruit, juice, and vegetables for fun and health: outcome evaluation [published erratum appears in Health Educ Behav 2000 Jun;27(3):390]. Health Educ Behav. 2000, 27 (1): 96-111.View ArticlePubMedGoogle Scholar
- Reynolds KD, Franklin FA, Binkley D, Raczynski JM, Harrington KF, Kirk KA, Person S: Increasing the fruit and vegetable consumption of fourth-graders: results from the high 5 project. Prev Med. 2000, 30 (4): 309-319. 10.1006/pmed.1999.0630.View ArticlePubMedGoogle Scholar
- Haire-Joshu D, Nanney MS: Prevention of overweight and obesity in children: Influences on the Food Environment. The Diabetes Educator. 2002, 28 (3): 415-422.View ArticlePubMedGoogle Scholar
- Glanz K, Lewis FM, Rimer BK: Health behavior and health education: theory, research, and practice. 1997, San Francisco , Jossey-Bass, 2Google Scholar
- Green LW, Kreuter MW: Behavioral and environmental assessment. Health Promotion and Planning. 1999, California , Mayfield Publishing Company, 145-3rd.Google Scholar
- Akan GE, Grilo CM: Sociocultural influences on eating attitudes and behaviors, body image, and psychological functioning: a comparison of African American, Asian American, and Caucasian college women. International Journal of Eating Disorders. 1995, 18 (2): 181-187.View ArticlePubMedGoogle Scholar
- Bronner Y: Nutritional status outcomes for children: Ethnic, cultural, and environmental contexts. Journal of the American Dietetic Association. 1996, 96 (9): 891-903. 10.1016/S0002-8223(96)00242-8.View ArticlePubMedGoogle Scholar
- Thompson B, Demark-Wahnefried WD, Taylor G: Baseline fruit and vegetable intake among adults in seven 5 a day study centers located in diverse geographic areas. J Am Diet Assoc. 1999, 99: 1241-1248. 10.1016/S0002-8223(99)00306-5.View ArticlePubMedGoogle Scholar
- Xie B, Gilliland FD, Li YF, Rockett HR: Effects of Ethnicity, Family Income, and Education on Dietary Intake among Adolescents. Preventive Medicine. 2003, 36: 30-40. 10.1006/pmed.2002.1131.View ArticlePubMedGoogle Scholar
- DiSogra L, Glanz K: The 5 A Day Virtual Classroom: an on-line strategy to promote healthful eating. Journal of the American Dietetic Association. 2000, 100 (3): 349-352. 10.1016/S0002-8223(00)00106-1.View ArticlePubMedGoogle Scholar
- Kumanyika S, Tell GS, Shemanski L, Polak J, Savage PJ: Eating patterns of community-dwelling older adults: the Cardiovascular Health Study. 4. 1994, Ann Epidemiol-Google Scholar
- Shi L: Sociodemographic characteristics and individual health behaviors. Southern Medical Journal. 1998, 91 (10): 933-941.View ArticlePubMedGoogle Scholar
- Kirby S, Baranowski J, Reynolds K, Taylor G, Binkley D: Children's fruit and vegetable intake: Sociodemographic, adult child, regional, and urban-rural influences. Journal of Nutrition Education. 1995, 27: 261-271.View ArticleGoogle Scholar
- Hearn MD, Baranowski J, Doyle C, Smith M, Lin LS, Resnicow K: Environmental influences on dietary behavior among children: availability and accessibility of fruits and vegetables enable consumption. J Health Educ. 1998, 29: 26-32.View ArticleGoogle Scholar
- Kratt P, Reynolds K, Shewchuk R: The Role of Availability as a moderator of Family Fruit and Vegetable Consumption. Health Education and Behavior. 2000, 27 (4): 471-482.View ArticlePubMedGoogle Scholar
- Morland K, Wing S, Roux AD: The Contextual Effect of the Local Food Environment of Residents' Diets: The Atherosclerosis Risk in Communities Study. American Journal of Public Health. 2002, 92 (11): 1761-1767.View ArticlePubMedPubMed CentralGoogle Scholar
- Buzzard M: 24-hour dietary recall and food record methods. Nutritional Epidemiology. Edited by: Willett W. 1998, New York , Oxford University Press, 2ndGoogle Scholar
- Continuing Survey of Food Intakes by Individuals (CSFII) and the Diet and Health Knowledge Survey (DHKS), 1994-96. 1998Google Scholar
- Serdula MK, Gillespie C, Kettel-Khan L, Farris R, Seymour J, Denny C: Trends in fruit and vegetable consumption among adults in the United States: Behavioral Risk Factor Surveillance System, 1994-2000. American Journal of Public Health. 2004, 94: 1014-1018.View ArticlePubMedPubMed CentralGoogle Scholar
- Nanney MS, Hessler K, Kelly C, Haire-Joshu D: Identifying usual portion sizes for commonly eaten foods among 7-9 year olds using parent-child consensus. 2003, Saint Louis University School of Public HealthGoogle Scholar
- Nielsen SJ, Popkin BM: Patterns and trends in food portion sizes, 1977-1998. American Medical Association. 2003, 289 (4): 450-453. 10.1001/jama.289.4.450.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.