Abstract:

Picky eating is a common dietary issue among children characterized by lack of variety of foods consumed due to rejection of familiar (or unfamiliar) foods. The influencing factor model of picky eating behavior in children indicates that environmental and cognitive factors are key elements influencing this. Studies have found that instrumental feeding exacerbates picky eating behavior in children. However, due to the relatively young age of children in previous studies, research on the relationship between instrumental feeding and picky eating behaviors in school-aged children is insufficient. Furthermore, the brain plays a central role in guiding eating behavior; however, to date, limited neuroscientific research on the neural basis of picky eating behaviors in school-aged children exists. This study aimed to utilize resting-state functional magnetic resonance imaging (rs-fMRI) data combined with a machine learning method to explore the neural basis of picky eating behaviors in children. Additionally, it attempted to show the neural mechanisms through which instrumental feeding influences picky eating behavior.

A total of 139 children were recruited for this study. Instrumental feeding and picky eating behaviors were assessed through parent-reported measurements and rs-fMRI was conducted. A total of 87 children were included in the formal analyses as those who did not participate in the two behavioral measurements and with unqualified rs-fMRI scans were excluded. This study utilized regional homogeneity and functional connectivity to evaluate the resting-state neural substrates of picky eating behaviors. Subsequently, a machine learning method is employed to validate the stability of our results. Additionally, a mediation model was constructed to investigate the mediating role of resting-state neural substrates in the relationship between instrumental feeding and picky eating behavior.

Results showed that picky eating behavior was positively correlated with regional homogeneity in the right caudate (see Figure 1, r = 0.43, p< 0.001, 95% CI = [0.25 0.59]). Functional connectivity results showed that picky eating behavior was positively correlated with functional connectivity between the right caudate and left putamen (see Figure 2, r = 0.43, p < 0.001, 95% CI = [0.24 0.59]). A prediction analysis based on a cross-validation machine learning method indicated a significant correlation between picky eating behavior scores predicted by the aforementioned neural substrates (i.e., regional homogeneity in the right caudate and functional connectivity between the right caudate and left putamen) and the actual observed picky eating behavior scores (regional homogeneity: r_{(predicted, observed)} = 0.37, p < 0.001; functional connectivity: r_{(predicted, observed)} = 0.35, p < 0.001). Shown as Figure 3, the mediation model further suggested that functional connectivity between the right caudate and left putamen could mediate the relationship between instrumental feeding and picky eating behavior (indirect effect: β = -0.16, standard error = 0.05, 95% CI = [-0.26 -0.06]).

Specifically, instrumental feeding might negatively influence the functional connectivity between the right caudate and left putamen, and further reduce picky eating behavior.

By combining resting-state regional homogeneity and functional connectivity analyses, this study detected altered functional brain activity related to picky eating behaviors in children aged 9 to 12. Specifically, hyperactive neural interactions within the brain areas involved in sensory sensitivity and reward processing may explain the manifestation of picky eating behavior in children. Additionally, instrumental feeding negatively influences picky eating behavior through brain activity in regions involved in sensory sensitivity and reward processing. This study provides new insights into the resting-state neural substrates of children's picky eating behavior, extends the influencing factor model of children's picky eating behavior, and provides theoretical support for interventions to improve poor picky eating behavior in children through parental feeding practices.

Key words: picky eating behavior, instrumental feeding, children, resting-state fMRI