A curriculum based on social learning theory emphasizing fruit exposure and positive parent child-feeding strategies: A pilot study

Methods 

Participants in this research study were recruited from seven elementary schools within the Princeton school district of Greater Cincinnati. Letters explaining the purpose of the study were mailed to the parents of all children in the fifth and sixth grades at the schools. Twenty-six parents and their children expressed an interest in participating in the study; nine pairs actually participated in at least eight of the 10 intervention classes, and 17 came to the first class (which consisted of an orientation and pre-evaluation session) but were unable to attend any follow-up classes. This latter group of 17 parent-child pairs served as the control group. This study was approved by the institution's internal review board.

The intervention was conducted on Saturday mornings in the food science lab of the Department of Nutritional Sciences at the University of Cincinnati. The curriculum was designed to engage students and their parents in learning experiences to increase their intake of a variety of whole fruits. Our rationale for choosing only one food group was that we wanted to allow an adequate amount of time for children to learn about and be exposed to only a few foods to ensure maximal program effectiveness. A nutrition graduate research assistant, with training in nutrition education, taught all children and parent classes. The Social Learning Theory ((10)) provided the theoretical basis for the intervention. Each session of the 10-week program was two hours in length. Lessons consisted of one hour of instruction provided to children and parents devoted to enhancing knowledge about a fruit, skills necessary to incorporate the fruit in the diet, and techniques to enhance goal-directed behaviors. After this hour, children remained in the food laboratory, where they completed simple pen-and-paper activities to reinforce the concepts learned in the first portion of the class. Parents were taken into a separate room, where they were given instruction on child-feeding strategies to enhance fruit acceptance. The sessions included interactive discussions, case studies, brainstorming activities, and games. Topics covered in the classes included exposure (increasing the availability of healthful foods), monitoring (understanding that the children can self-regulate the amount of food they need each day), restriction (learning how to present a variety of foods to children without restricting access to certain foods), rewarding/punishing (understanding that using food as a reward or punishment can be counterproductive), and encouragement (learning how to present food to a child in a nonpressured manner without using verbal prompting). In the final half-hour of the session, parents rejoined their child in the food laboratory. During this period, parents and children were given an opportunity to sample 10 different fruits (apple, orange, banana, grape, pear, star fruit, mango, papaya, kiwifruit, and cantaloupe). The same 10 fruits were used in each class to give children several opportunities to taste them. Also during this segment of the class, parents were encouraged to practice feeding strategies they had learned earlier.

Outcome measures assessed in this study were change in knowledge, preference, and intake of fruit for children, and change in control over child feeding for parents. Outcome evaluation instruments were administered in a pretest/posttest manner. Change was determined by subtracting pretest from posttest scores or intake.

The knowledge questionnaire tested curriculum-specific knowledge about fruits using 16 multiple-choice questions, with a possible scoring range of 0 to 16. The questionnaire was developed specifically for this study and went through several stages of pilot testing to establish appropriate literacy level.

The food preference questionnaire, consisting of a list of the 10 fruits targeted during the intervention, was designed to measure changes in preference for fruit. A 9-point hedonic rating scale measured degree of preference ((11)). Using this scale, the child was asked to rate how much he or she liked or disliked each fruit listed. No response was recorded if the child had never tried the food. The questionnaire had a possible scoring range of 10 to 90.

Whole fruit intake was determined by analyzing subjects' fruit intake before and after the study from food intake records. Children were trained by a registered dietitian on how to complete a three-day food record (two weekdays and one weekend day) and on the use of a two-dimensional food model to help determine serving sizes for foods eaten. Once completed, a registered dietitian reviewed each food record for accuracy. The fruit intake score was determined by manually counting each half-serving of whole fruits and 100% fruit juices.

Change in control over child feeding was measured using the Child Feeding Questionnaire developed by Birch and Johnson [12], [13]. Fifteen questions are included on the questionnaire, which assesses parental control over child feeding in the dimensions of restriction, encouragement, and monitoring. A five-point Likert scale for each question is used to measure individual differences in ideas about child feeding. Possible score ranges on each control dimension are as follows: restriction, 8 to 40; encouragement, 4 to 20; and monitoring, 3 to 15. Higher scores within the range for each dimension indicate a greater degree of parental control. A total control score, with a possible range of 15 to 75, is derived from summing the individual scores across dimensions of control.

Statistical analyses were performed using the Statistical Analysis System software (version 6.12, 1998, SAS Institute, Cary, NC). Descriptive statistics were used to summarize the demographic characteristics of the subjects. Categorical data were compared among the intervention and comparison groups using the χ2 procedure. Differences between the two groups for continuous variables (demographic characteristics and changes in outcome measures) were assessed using separate one-way analysis of variance.

Results and discussion 

Demographic characteristics of the child and parent groups are presented in Table 1.Differences were found between groups with respect to child gender and child and parent race. The child control group had a significantly greater number of females, and child and parent control groups had a greater number of white persons than did the intervention group. Consequently, gender (for children) and race (for children and parents) were used as covariates in subsequent statistical analyses.

Table 1. Subject demographic

	Characteristics	Intervention group	Control group
	N (child-parent pairs)	9	17
	Age (yr)	←Mean ± SD→
	Children	10.8±0.6	11.6±0.5
	Parents	41.7±3.7	42.3±1.9
	Sex (%)
	Children (males/females)*	66.7/33.3	17.7/82.3
	Parents (males/females)	11.1/88.9	0/100
	Race (%)* (child and parent race)
	White	55.5	94.1
	Other	44.5	5.9

*P<.05, as calculated by χ2 test. SD=standard deviation.

Comparison of differences among children in the intervention and control groups for change in fruit knowledge, preference, and intake are shown in Table 2.Analysis of variance of change scores for these variables showed a significant increase in knowledge scores and daily whole fruit intake after intervention vs before intervention in the experimental group as compared with the control group. However, there were no differences between the two groups with respect to fruit preference scores.

Table 2. Changes in children's fruit knowledge, preference, and intake and changes in parental control over child-feeding: Intervention participants vs control

		Intervention group (n=9)			Control group (n=17)
		Pre-test	Post-test	Change	Pre-test	Post-test	Change	P value
		←Mean±SD→			←Mean±SD→
	Fruit score
	Knowledgea	10.2±1.8	13.7±1.7	3.5±1.7	12.3±1.2	12.4±1.2	0.1±1.1	.001*
	Preferenceb	72.2±6.6	72.1±12.6	−0.1±6.6	69.8±10.4	68.5±5.3	−1.2±7.2	NS
	Intakec	1.8±2.3	2.6±2.3	0.7±2.2	1.0±1.0	0.8±0.6	−0.2±0.9	.02*
	Control typed
	Restriction	22.7±8.3	20.6±9.5	−2.1±5.2	27.2±4.8	30.4±3.3	3.2±4.0	.02*
	Encouragement	8.9±3.5	5.9±2.0	−3.0±3.5	11.3±4.9	8.8±3.5	−2.5±5.0	NS
	Monitoring	12.0±3.1	9.7±4.2	−2.3±3.2	10.6±2.4	12.2±4.5	1.6±3.7	.03*
	Total control	43.2±9.3	35.7±11.5	−7.4±8.3	46.5±7.1	51.7±7.0	5.2±3.9	.01*

aFruit knowledge score range: 0-16, an increase in score indicates gain in knowledge. bFruit preference score range: 10-90, an increase in score indicates an increase in fruit liking. cCounted in serving portions. dPossible score ranges: Restriction 8-40; Encouragement 4-20; Monitoring 3-15; Total Control 15-75; a lowering of score indicates a reduction in parental control. *Significant at P < .05. SD=standard deviation. NS=not significant.

Table 2 also shows differences among parents in the intervention and control groups for degree of parental control over child feeding in the dimensions of restriction, encouragement, and monitoring. Analysis of variance showed significantly lower change scores for restriction and monitoring of child feeding in the intervention group as compared with the control group, indicating that parental control in these areas was lower after vs before intervention. There was no difference between the two groups with respect to change scores in the area of encouragement. Overall, parents who participated in the intervention reported using less control over child feeding after vs before intervention as compared with the control group.

Our findings suggest that the curriculum elements included in this intervention effectively promoted change in dietary behavior (fruit intake) in children and may have potential for changing intake of other food groups shown to be inadequate in the American diet. Reasons for the lack of change in fruit preference among intervention participants are unclear. Typically, changes in food preference are noted before changes in food intake ((7)). Baseline fruit scores for both groups were relatively high, thus leaving little opportunity for improvement.

School-based dietary intervention programs often only minimally involve parents [14], [15]. Generally, a newsletter is sent home to parents informing them of nutrition-related classroom activities. In this intervention, parents were active participants in classroom learning, skill building activities, and fruit tasting. Additionally, they were taught positive ways to present foods to their child to promote food acceptance. Because parents' eatingbehaviors, attitudes, and child-feeding practices have a major impact on the development of children's food acceptance and intake patterns ((16)), including them in nutrition intervention activities in the schools would be ideal. Previous research has shown that increased parental control over child feeding may inadvertently result in reducing a child's preference and intake of healthful foods [4], [6], [7]. This study suggests that school-based dietary interventions including parents and parent-focused child-feeding protocols may effectively increase acceptance and intake of health-promoting foods by children. A worthwhile follow-up to the present study would be to conduct a study that could specifically measure the individual effects of the different components of this intervention (parent vs child components) to determine which section best predicted the final positive outcomes.

Application 

■This study showed that a nutrition intervention program consisting of extensive food exposure and parent-focused lessons on child-feeding strategies is feasible and can be effective. Nutrition professionals must acquire the skills to develop interventions that integrate both food lessons and repeated food tasting for children along with feeding protocols for parents to effectively promote acceptance and intake of healthful foods in children.