To address childhood malnutrition, the use of multifaceted methodologies, such as mixed methods research, is required to inform effective and contextual interventions. However, this remains limited in studying malnutrition among school children in a South African context, notwithstanding its persistence. We adopted a convergent parallel mixed methods design to best understand the magnitude of malnutrition through multilevel influences in a rural area. A quantitative survey determined the magnitude of malnutrition and associated factors among school children and their mothers (n = 508), parallel to a qualitative study, which explored mothers' insights into the influences of child growth and nutrition in interviews using seven focus group discussions. Mixed methods integration was achieved through convergence of the quantitative constructs developed from measured variables for malnutrition and related factors with ten emergent qualitative themes using a joint display analysis to compare the findings and generate meta-inferences. Qualitative themes on food unavailability and affordability, poor feeding beliefs and practices, and decision to purchase foods were consistent with the quantified poor socio-demographic status of mothers. Furthermore, the qualitative data explained the high prevalence of undernutrition among children but did not corroborate the high estimated households' food security in the quantitative survey. The misperceptions of mothers on child growth agreed with limited food knowledge as well as lack of knowledge on child growth gathered during the survey. Moreover, mothers believed that their children were growing well despite the high presence of childhood undernutrition. Mothers further overrated the effectiveness of school feeding programmes in providing healthy food to children as compared to their household food. They reported high incidence of food allergies, diarrhea, and vomiting caused by food consumed at school which resulted in children not eating certain foods. This might have impacted on the nutritional status of children since mothers depended on the school feeding program to provide food for their children. The ambiguity of cultural influences in relation to child growth was evident and substantiated during qualitative interview. Mixed methods integration offered a better understanding of malnutrition from empirical findings on interrelated factors at child, maternal, household, and school levels. This study points to a need for multilevel, informed, and contextual multidimensional interventions to contribute towards addressing childhood malnutrition in South Africa.
Keywords: South Africa; malnutrition; mixed methods; mothers; multilevel factors; rural context; school children.