Effectiveness of an IMB-based Intervention for Reducing Sweetened Beverages Consumption in Preschool Children

BACKGROUND. Sugar sweetened beverages (SSB), including sodas, fruit drinks, nectars, and ice teas, make up the largest source of added sugar in the American diet. Fruit juices and fruit drinks alone contribute 17-18% of the energy in children's diets. For young children, beverage consumption depends on beverages available and served at home by parents. Consequently, the most effective pathway for reducing SSBs in pre-school children must target the parent as the agent of change. The IMB model is the theoretical foundation for behavior change in this project. Unlike other behavior change models, IMB incorporates three major constructs that theoretically interact, leading to the outcome behavior. Each construct of IMB (information, motivation, and behavioral skills) includes subconstructs as a basis for tailoring the model to an individual health-related behavior.

RESEARCH DESIGN AND METHODS. The proposed design allows for 2 sets of systematic comparisons that will address the major study hypotheses: 1) analysis of the differences in outcomes between intervention and control groups at baseline, 1-week post-intervention and 6 months post-intervention; and 2) analysis of the IMB factors that mediate the impact of the intervention on the key outcome measures.

Participants in the intervention will receive 10 weeks of SSB-related nutrition education at their child's preschool or daycare. Twice a week, two student educators will be available to parents (both research participants and casual observers) at each site for 2 hours during the typical afternoon pick-up. Each week a different lesson pertaining to parental information, motivation, or behavioral skills related to SSB will be presented via an interactive display board. Lessons will last 5 to 10 minutes, as parents realistically only have this amount of time to stop at an educational board before or after picking up their child. These displays will include an activity for parents to apply knowledge and skills learned during the short lesson. Educators will also provide parents with informational handouts pertaining to each lesson to reinforce concepts learned. In addition, the program will include educational flyers and posters for site promotion. All materials to be developed will be culturally sensitive, available in English and Spanish, and appropriate for this low-literacy population.

Target Population. Twenty Hartford area center or school based early childcare centers serving primarily low-income, minority families will be recruited to participate in the study.

Sample Size. Investigators estimated a target sample size of 400 to provide enough stability in the parameter estimates and power to detect effects in the major set of analysis for the current research. Based on our previous efforts in conducting preschool based prevention research, the investigators expect follow-up retention rates to be 75%. Families who move or leave the site during the study will be retained for analysis.

Recruitment. Two weeks prior to baseline data collection, researchers will begin participant recruitment. The investigators have experience with recruitment in preschools. The researchers will obtain parent written consent and child verbal assent prior to individual involvement in the study.

Data Collection. Procedures used to collect data from parents and children will be identical across all 3 time points: baseline (1 week prior to intervention), 1-week post-intervention, and 6 months post-intervention. At each point, two periods of data collection will occur: 1) at a prearranged meeting at the child's day care, researchers will take anthropometric measurements, administer the IMB, Parent Attitude Toward Nutrition and Child Health, and Food Security questionnaires, and train the parent on the food record and HBI; 2) on the following Monday, the researcher will call the parent to collect data from the HBI and conduct the 48 hour food recall aided by the food record. Since the HBI and food recall will be completed over the phone, participants must have access to a telephone. The 1 week and 6-month follow-up assessments will repeat collection of all data gathered at baseline.

Instruments and Measures

1. IMB Survey. A structured researcher-administered survey instrument will be used to assess parental information, motivation, and behavioral skills related to their child's SSB and demographic characteristics. A sum of construct items will create the final score for the three construct indices (Information, Motivation, and Behavioral Skills).
2. Home Beverage Inventory. Using the HBI journal as a guide, participants will record all non-alcoholic beverages in their home, including liquids, powders, concentrates, teas, coffees, sweeteners, and syrups. They will look for beverages throughout their home but will be reminded to specifically look in their freezer, refrigerator, cabinets, pantries, cupboards, counters, and closets. Participants will report the beverage brand, type, flavor, size, container, and number of containers for each drink. The interviewer will record the HBI, as reported by participants, while using probes to ensure clarity and completeness. Analysts will use data collected from the HBI to calculate the availability of sugar sweetened beverages in the home via numbers of containers and number of types of beverages available.
3. Weekend Food Recall. The investigators will collect food and beverage consumption data for two weekend days when the parent will be controlling the food supply and aware of their child's full-day intake and will use a multi-pass food recall aided by a food record. A researcher will phone each parent on Monday to conduct the food recall using a modified United States Department of Agriculture (USDA) multi-pass protocol
4. Parental Attitudes Toward Nutrition and Child Health Questionnaire. This interviewer-administered questionnaire will be composed of nutrition and health related constructs whose items have demonstrated empirical validity in our study population, such as parent perceived feeding responsibility and monitoring of food intake from the modified version of the Child Feeding Questionnaire and parent feeding control from the Parental Feeding Style Questionnaire.
5. Food Security Questionnaire. The researchers will measure household food security using the validated USDA Food Security Module which consists of 18 questions that ask with increasing severity about a household's experiences with getting enough food.
6. Anthropometric Measurements. Parent and child heights will be measured in stocking feet using a stadiometer using standard protocols. Parent and child weights will be obtained using an electronic self-calibrating digital scale using standard procedures. Body mass index (BMI) will be used as a measure of adiposity in parents and children.

Data Analysis. Data analysis plans for the research questions listed on page 1 and 2 are presented below. Because this study utilizes a randomized multisite design in which participants are nested within daycare sites, scientific tests of the effectiveness of the intervention will be performed in a series of regression models using SUDAAN, a proprietary statistical package developed by the Research Triangle Institute. SUDAAN was developed to address generic problems in the analysis of hierarchical data structures, that is, data in which characteristics of one unit of analysis (e.g., individuals) are nested within, and vary among, larger units (e.g., social groups or contexts).

Descriptive Analysis. Before substantive issues regarding the efficacy of the SSB intervention are examined, the evaluators will perform analyses to determine the comparability of the intervention and control groups. This will enable us to determine the adequacy of the randomization in terms of demographic characteristics and outcomes of interest. The researchers will perform one-way ANOVAs (for continuous baseline characteristics) and chi-square tests (for categorical baseline characteristics) in SUDAAN to explore differences among the 2 groups. Subsequent analyses will statistically adjust for any differences in the composition of the intervention and control groups.