Child Feeding Behaviours, Nutritional Status and Energy Intake Regulation among Urban Ghanaian Children

Abstract

Background: The recent Ghana Demographic Health Survey indicates that infant malnutrition still perpetuates in Ghana. Energy regulation, which is the “internal driving force for search, choice, and ingestion of food”, plays a critical role in the causation, prevention, and treatment of malnutrition. Children have an inherent ability to self-regulate energy intake (EI). Successful energy regulation is essential in maintaining a healthy body weight and overall wellness. It is not clear whether undernourished children regulate energy intake similarly to healthy children. This information is important, especially in settings where many children are undernourished. There is also a modicum of knowledge of how energy regulation develops in children and its effect on their nutritional status. Aim: This study aimed to describe the extent to which undernourished children (with moderate acute malnutrition (MAM) and/or stunted) living in urban Ghana are able to regulate their energy intake and to determine whether genetic, behavioural and environmental factors predict this outcome. Methodology: This was a two-phase study: a cross sectional study (n=262) and an experimental study (n=41) carried out among child-caregiver dyads (child 1-3 years old) at the Ashiedu Keteke Municipal district in Greater Accra Region-Ghana. The cross-sectional study carried out at selected child welfare clinics involved parental/ caregiver rating of children’s dietary intake and eating behaviour using the International Complementary Feeding Evaluation Tool (ICFET). The questionnaire also assessed child’s illness, household hunger and socioeconomic factors such housing characteristics, water and sanitation facilities. Saliva samples were collected from children for genetic testing for polymorphisms in genes associated with energy regulation. Nine single nucleotide polymorphisms (SNPS) were genotyped, and a polygenic risk score (PGRS) was generated. The experimental study was a within-subject crossover preload experiment conducted at the homes of randomly selected children from phase 1 of the study. A standardized satiation methodology was employed to assess the ability to compensate for a drink taken before a meal. For the experiment, caregiver-child dyads were visited twice in their homes. Children were randomly assigned to two conditions: high energy preload drink supplying 10% of daily energy requirements or a low energy drink supplying 9kcal/100ml, after at least 2 hours fast. They were served identical ad libitum lunches 30 minutes after consuming the preload. Calories consumed from preload were calculated and calories from lunch meals were analysed using bomb calorimetry. The proportion of the preload compensated (COMPX) was calculated. The weight and height/recumbent length of all the children in both studies, were measured following standard protocols and converted to Z-scores using WHO Anthro and Anthro Plus software. Data was entered and analysed using SPSS software version 22. Chi square test was used to test associations between categorical variables. T-test and one-way ANOVA were used to compare means across groups. Linear regression and multinomial regression were used to test predictive associations of energy regulation and nutritional status. A p-value <0.05 at a 95% confidence level was considered as threshold for statistical significance. Results: A total of 262 child-caregiver pairs (average age of child 19.25 ± 6.65 months) participated in the cross-sectional study, of which 75 were wasted (MAM), 53 were stunted, and 134 were classified as having normal anthropometry. There was a significant negative correlation between appetite and food refusal (r= -0.746, p<0.001), forced-feeding (r = -0.382, p<0.001) and caregiver feeding anxiety (r = -0.698, p<0.001). A positive correlation was seen between food refusal and forced-feeding (r= 0.351, p<0.001), food refusal and caregiver feeding anxiety (r= 0.624, p <0.001), as well as forced-feeding and caregiver feeding anxiety (r= 0.297, p<0.001). Wasted children had a significantly lower appetite z-score (-0.27 ± 1.1), a higher z-score for food refusal (0.22 ± 1.1) and a higher z-score for caregiver feeding anxiety (0.49 ± 1.0) compared to both stunted and healthy children (p<0.05). There were no significant differences in feeding behaviour between the stunted and normal groups. Feeding anxiety was a consistently significant predictor of nutritional status after adjusting for confounders (p<0.001). The SNP rs2274333, situated near the CA6 gene, exhibited a higher frequency of homozygosity for the effect allele, AA, in children with wasting compared to their healthy counterparts. There was a significant association between PGRS and appetite (p=0.046). There was no significant association between PGRS and nutritional status. There were 41 children (17 MAM, 12 stunted and 12 healthy) in the experimental study. The overall average compensation was 62.43%. More wasted children had either minimal or no compensation (41.2%) or over-compensation (35.3%), and the majority (66.7%) of the stunted children had minimal or no compensation, and none over-compensated: this difference was statistically significant (p=0.023). Conclusion: This study has given insight into the complex interaction between genetics, environmental factors, and nutrition among children. Although a longitudinal study would be more suitable for ascertaining causal directions, the findings from this study have implications for research, practice, and policy, especially regarding appetite, feeding behaviour and attitudes of caregivers. Efforts to prevent or treat undernutrition in children might benefit from focusing on modifiable factors such as dietary factors and feeding behaviours of both children and their caregivers. This also holds promise for precision nutrition as future and larger studies accentuate the genetics of undernutrition.