Systematically testing the effects of promotion techniques on children’s fruit and vegetables intake on the long term: a protocol study of a multicenter randomized controlled trial

Recently an overarching theoretical model has been developed through an eclectic synthesis of existing theoretical models from different disciplines and recent empirical evidence has been conducted that explains and predicts whether, how, when, and for whom food promotion techniques increase children’s fruit and vegetables intake, both on the short- and long-term [40]. The four foundational assumptions of the theoretical model, The Promotion of Fruit and Vegetables Model are that (1) by increasing the reinforcing value of fruit and vegetables (e.g., liking and wanting) through effective food promotion techniques, (2) a reciprocal relation with eating behavior occurs, that in time (3) leads to a normalization of intake of fruit and vegetables (habitual formation). Furthermore, (4) individual and contextual factors (e.g., BMI [41, 42], SES [12, 13], Food Fussiness [43, 44], Parental Feeding Style [45, 46]) determine individual susceptibility to food promotion. In the next section, I will explain the model and associated hypotheses.

Human eating behavior is not only guided by a conscious and reflective system, but also via a non-conscious, impulsive, associative system that is susceptible to food promotion techniques [20, 46‐50]. By using different food promotion interventions within the vlogs, like using unboxing, preparing and taste testing the fruits and vegetables (taste) versus using a challenge (humor and fun) [51] to increase the rewarding value of fruits and vegetables, it will be possible to test differences between media messages [19, 20, 52‐67]. The current project proposes that food promotion techniques increase the reinforcing value of fruits and vegetables among young children, which in turn has a positive effect on the immediate intake.

Most studies have examined immediate or short-term effects of healthy food promotion techniques [65‐68], while long-term effects are particularly important for vital health improvements [1‐9]. Numerous studies have shown that repeatedly exposing and tasting fruit and vegetables increases the liking of it, thereby increasing the probability of consumption [69‐74]. Availability and mere exposure are processes by which experiences of a stimulus positively enhances the attitudes towards it, resulting in the given stimulus acquiring positive valence and becoming more apparent in the mental scheme of children [69, 75], thereby habitualizing the consumption of fruit and vegetables [74, 75] and possibly even substituting the HFSS foods by healthier snacks [53, 74‐78].

Multiple individual and contextual factors have been established that explain susceptibility to food promotion of unhealthy food [20, 24]. For example, age [42, 43], impulsivity [79], weight status [42, 43] and attentional bias [80] have been found to moderate the effect of unhealthy food promotion on children’s intake, but I will investigate whether these concepts moderate the effect of promotion of fruit and vegetables. Other individual factors that are distinguished as possible moderating factors are neophobia and food fussiness [44, 45]. Additionally, contextual factors, like parental feeding techniques [46, 47] or social economic status [81, 82] are highly related to the intake of fruit and vegetables. Establishing individual and contextual dispositional factors, are vital for our understanding the variability in the processing of the promotion of fruit and vegetables, the reinforcing value of these foods, and subsequent intake.

First, to develop the vlogs in co-creation different focus groups will be held with (1) children (N = 25, between 8 and 13 years), (2) parents (N = 10), (3) vloggers (N = 5), and (4) fruit and vegetable producers and marketers (N = 5). The main question for the focus groups are: “What should be the exact content and methods of the vlog to increase (which) fruits and vegetables among the target group?”. The intervention materials will be pilot-tested among the target group (N = 100). The primary outcomes from these focus groups will be used to develop the exact intervention.

Second, participating primary schools (from low-, middle- and high-SES areas) in the randomized controlled experimental intervention-study (computer-generated randomization) will receive a food stand located on a key-point in school, with standard fresh, free, and ready-to-eat fruit and vegetables that can be consumed during the breaks [63]. The number of children that consume fruit and vegetables will be counted (with a clicker [63]) and calculated afterwards by the experimenter and research assistant. Children in the experimental condition will watch the promotional vlog (10–20 min) three times a week. A mixed repeated measure design with three different conditions will be used: (1) control, (2) a vlog unboxing fruit and vegetables (preparing and tasting), and (3) a vlog doing a challenge with the fruit and vegetables (e.g., contests, tricks, games). Written consent from the principal and teachers will be obtained first, followed by written consent from parents and verbal consent from children. Children can stop participation during the experiment without any reason. Collected data from this participants will be deleted. All collected data will be pseudonymized, whereas anonymized data will be made available for other researchers. Randomization will take place at school level.

Power analyses (G*Power) [83] suggest that N = 342 is expected to be sufficient to detect significant (α = 0.05), medium effects (f = 0.25) of conditions (power = 0.80). Next, children will complete questionnaires including pre- and post-test measurements during school hours together with the experimenter to assess general dietary intake [13], before (baseline, T1) and immediately after the intervention (T2, 10-weeks after the baseline, see Fig. 1). Primary outcome of this study is the intake of fruit and vegetables. Secondary outcomes are liking and wanting of fruit and vegetables [28], attitudes towards the vlog [79, 80], attitudes towards the food-stand [79, 80], attitudes towards and intention to consume different foods (e.g., fruit, vegetables, HFSS foods) [79, 80], advertising defenses [79, 80], brand and product recall and recognition [79, 80]. These outcomes will be assessed via questionnaires together with the experimenter. Children’s age, hunger [79, 80], sex, SES [12, 13], and BMI [79, 80] will be included as control or moderating variables. All measures have been validated among this age group in previous research.

After 6 (T3) and 12 months (T4) follow-up measurements will take place. Actual intake during lunch in schools, liking of fruit and vegetables, and attitudes towards and intention to consume foods, will be assessed again during school time.

Bayesian mixed effects models will be carried out in R using the brms package [84] to examine differences in all dependent and mediating constructs between pre- and post-test conditions. In addition, multilevel regression analyses will be conducted to examine possible group differences. Structural Equation Modeling will be used to examine whether the liking and wanting of fruits and vegetables mediates the effect of the promotion of fruit and vegetables on the intake.