| dc.description.abstract | Background
The healthiness of children’s diets, and the success of interventions seeking to improve them, are hindered by children’s preferences for foods high in fat, salt and/or sugar over healthier options such as fruit and vegetables. Children’s food choices are often driven by hedonic factors such as taste rather than considerations of health and nutrition, meaning that educational approaches are unlikely to be successful on their own. Food Go/No-Go training (which sits in the family of motor response training and inhibition training tasks) is a computerised intervention that appears to target the automatic processes driving food choice (e.g., food liking/reward responses, and automatic motor responses) by requiring participants to inhibit their responses to certain foods in the context of a reaction time game. Chapter One presents an overview of the food Go/No-Go training literature.
Aims
In a series of experimental and feasibility studies, we aimed to answer the following questions: (i) can food Go/No-Go training lead to healthier food choices among children aged 4-11 years, (ii) how do variations in food Go/No-Go training tasks (e.g., different response signals, different delivery methods) impact the effectiveness of training on children’s food choices in relatively controlled settings, and (iii) how acceptable and feasible are different delivery methods of food Go/No-Go training for implementing in real-world environments?
Methods
Chapter Two presents four early experimental studies in a school setting, comparing the effects of food Go/No-Go training against a control task on children’s food choices in a hypothetical choice task. Chapter Three presents an experimental study (also in a school setting) exploring whether using evaluative response signals (happy and sad faces) enhances training effects on children’s food choices compared to neutral signals (green and red symbols). Chapter Four presents two feasibility studies exploring the delivery of computer-based Go/No-Go training to families via the internet. Chapter Five presents an experimental study in a school setting, comparing computer-based training and touchscreen-compatible app-based training against a control task. Chapter Six explores the feasibility of delivering app-based Go/No-Go training to families in a mixed-methods feasibility study taking a randomised controlled design. Finally, Chapter Seven presents a qualitative study, using a thematic analysis method to explore the experiences of parents in the UK who were engaged in family healthy eating efforts during the coronavirus pandemic. Chapter Eight summarises the research presented here, presents the results of mini meta-analyses of the studies included in this thesis, and suggests avenues for future research.
Results
In Chapter Two, children selected a significantly higher number of healthy foods after playing food Go/No-Go training compared to two different control tasks. In Chapter Three, no effect of food Go/No-Go training was observed (regardless of whether response signals were evaluative or neutral) which may have been due to large group sizes during testing (e.g., higher distraction and social influence during food choice). Attrition was high in the feasibility studies of Chapter Four, and feedback from parents suggested that the training would benefit from gamification and adaptation for touch-screens. In Chapter Five, computer-based (but not touch-screen app-based food Go/No-Go training) led to children choosing significantly higher numbers of healthy foods versus control - low statistical power may have masked an effect of app-based training. The feasibility study in Chapter Six met the majority of continuation criteria for feasibility and acceptability of the intervention and methods, however some families engaged very little with the food Go/No-Go training, and parent feedback indicated a number of ways the training could be optimised to enhance engagement. Chapter Seven revealed a number of challenges faced by parents during the UK lockdown in March 2020 (e.g., increased influence of children’s food requests on feeding decisions), reiterating why accessible interventions directly targeting children’s food preferences are needed in the wider toolkit of family-targeted healthy eating interventions. The mini meta-analyses in Chapter Eight suggested that both computer-based and app-based food Go/No-Go training significantly impact children’s food choices, but that the effect size for computer-based training is twice as large as that for app-based training, in line with research with this specific app with adult samples.
Conclusions
Overall, these studies indicate that food Go/No-Go training is a feasible, acceptable and effective tool that could be used to encourage healthier food choices among children. Further research should be undertaken to understand why the app-based training used in these studies yielded a smaller effect size than computer-based training, and whether an optimised, app-based intervention can be developed that both engages children and encourages them towards healthier food choices. | en_GB |
