Diet and development among children aged 36-59 months in low-income countries

Abstract

Objective: To assess the associations between diet, stimulation and development among children 36-59 months of age in low-income and middle-income countries (LMICs).

Design: We pooled Demographic and Health Survey data on 12 126 children aged 36-59 months from 15 LMICs. Child diet indicators included dietary diversity score (DDS, range 0-7), minimum dietary diversity (MDD, defined as DDS ≥4) and animal source foods (ASFs) consumption. Child development was assessed using the Early Childhood Development Index and stimulation by the number of stimulation activities (range 0-6). Associations were assessed using generalised linear models.

Results: In our sample, 18% of children met MDD and 50% received ≥4 stimulation activities. The prevalence of suboptimal cognitive, socioemotional, literacy-numeracy and physical development was 24%, 32%, 87% and 11%, respectively. Higher DDS, meeting MDD and consuming ASFs were associated with 8%-13% more stimulation activities. Children who met MDD were slightly less likely to have suboptimal literacy-numeracy development compared with children who did not meet MDD: relative risk 0.97 (95% CI 0.95 to 1.00). DDS, meeting MDD and ASFs consumption were not associated with cognitive, socioemotional or physical development. However, there was evidence of positive associations between MDD and cognitive and literacy-numeracy development among subgroups of children, including those who received ≥4 stimulation activities or attended an early childhood care and education programme.

Conclusions: Child diet was associated with more stimulation activities. However, independent of stimulation, socioeconomic status and other factors, child diet appeared to be a prominent determinant only of literacy-numeracy development among children 36-59 months of age.

Keywords: child development; child health; global health.

Figure 1

Heterogeneity of the association between child minimum dietary diversity and suboptimal child development by child, maternal and household factors, comparing children who met minimum dietary diversity and children who did not. Displayed are only factors that modified the associations with at least one child development domain, that is, p value for interaction was <0.05. All values are relative risk and 95% CIs. All models applied country-specific cluster variables and sampling weights. Estimates controlled for household wealth, rurality, household size, access to improved sanitation and access to improved water source; maternal age, education and marital status; stimulation; child age, sex and attendance of an early childhood education programme (preschool) and country and survey year.

Figure 2

Heterogeneity of the association between child consumption of animal source foods and suboptimal child development by child, maternal and household factors, comparing children who consumed animal source foods and children who did not. Displayed are only factors that modified the associations with at least one child development domain, that is, p value for interaction was <0.05. All values are relative risk and 95% CIs. All models applied country-specific cluster variables and sampling weights. Estimates controlled for household wealth, rurality, household size, access to improved sanitation and access to improved water source; maternal age, education and marital status; stimulation; child age, sex and attendance of an early childhood education programme and country and survey year.